| *** brian_sc has joined #openstack | 00:00 | |
| notmyname | hsom: file a bug, I guess (sorry, it's getting late here and I don't feel well) | 00:00 |
|---|---|---|
| *** bengrue has quit IRC | 00:01 | |
| hsom | no worries, thanks for trying | 00:02 |
| vidd | Kiall, you still here? | 00:02 |
| *** vernhart has quit IRC | 00:03 | |
| *** hingo has quit IRC | 00:03 | |
| *** rods has quit IRC | 00:04 | |
| daMaestro | notmyname, there are debs and other information in that tgz | 00:05 |
| daMaestro | notmyname, for me to be able to stick it into a package and not ship extra data... i need a new tgz | 00:05 |
| *** haji has joined #openstack | 00:06 | |
| vidd | daMaestro, can you download the current tar.gz, rip it open, strip what you personally dont need and zip it back up? | 00:07 |
| *** dwcramer has quit IRC | 00:13 | |
| daMaestro | yes, but policy dictates we have an upstream file location that can be verified | 00:14 |
| daMaestro | i can put anything into the spec, it's just "proper" to have an upstream location to point to for source verification | 00:14 |
| daMaestro | for now, i've just made the spec unpack the upstream source and then unpack the inner source | 00:14 |
| *** afm1 has joined #openstack | 00:16 | |
| *** vernhart has joined #openstack | 00:16 | |
| *** afm has quit IRC | 00:17 | |
| *** vladimir3p has quit IRC | 00:20 | |
| *** vladimir3p has joined #openstack | 00:20 | |
| *** rnirmal has quit IRC | 00:23 | |
| *** misheska has joined #openstack | 00:26 | |
| *** vernhart1 has joined #openstack | 00:27 | |
| *** haji has quit IRC | 00:28 | |
| *** dwcramer has joined #openstack | 00:30 | |
| *** vernhart has quit IRC | 00:31 | |
| *** miclorb_ has quit IRC | 00:33 | |
| *** cdub has quit IRC | 00:37 | |
| *** haji has joined #openstack | 00:37 | |
| *** rsampaio has joined #openstack | 00:37 | |
| *** bengrue has joined #openstack | 00:39 | |
| *** miclorb_ has joined #openstack | 00:40 | |
| *** rsampaio has quit IRC | 00:42 | |
| *** afm1 has quit IRC | 00:48 | |
| *** rsampaio has joined #openstack | 00:49 | |
| *** livemoon has joined #openstack | 00:50 | |
| livemoon | morning | 00:51 |
| andrewbogott | What's the story with Horizon and gerrit? I'm following http://wiki.openstack.org/GerritWorkflow but dead end with "The current project name, openstack/horizon.git, is not a known project." | 00:51 |
| vidd | hello livemoon | 00:51 |
| *** dragondm has quit IRC | 00:52 | |
| *** nati2__ has quit IRC | 00:55 | |
| *** KumarKR has quit IRC | 00:56 | |
| *** pixelbeat has quit IRC | 00:58 | |
| *** bengrue has quit IRC | 00:59 | |
| *** bwong has quit IRC | 01:00 | |
| *** jdg has quit IRC | 01:03 | |
| *** sdake has joined #openstack | 01:06 | |
| *** cdub has joined #openstack | 01:08 | |
| *** Fu4ny has joined #openstack | 01:08 | |
| *** daMaestro has quit IRC | 01:10 | |
| *** jog0 has quit IRC | 01:12 | |
| *** nati2 has joined #openstack | 01:15 | |
| *** dendrobates is now known as dendro-afk | 01:17 | |
| *** sandywalsh has quit IRC | 01:17 | |
| *** maplebed has joined #openstack | 01:18 | |
| *** rnorwood has quit IRC | 01:18 | |
| maplebed | hi, I'm going through http://swift.openstack.org/howto_installmultinode.html and am adding devices to the rings. | 01:18 |
| *** Fu4ny has quit IRC | 01:18 | |
| maplebed | the example says swift-ring-builder account.builder add z$ZONE-$STORAGE_LOCAL_NET_IP:6002/$DEVICE $WEIGHT | 01:18 |
| maplebed | does it require an IP address or can it use a hostname instead? | 01:18 |
| *** miclorb__ has joined #openstack | 01:19 | |
| *** miclorb_ has quit IRC | 01:19 | |
| *** obino has quit IRC | 01:22 | |
| maplebed | (for the $STORAGE_LOCAL_NET_IP field, that is) | 01:22 |
| *** jj0hns0n has joined #openstack | 01:22 | |
| *** mnour has joined #openstack | 01:22 | |
| *** sdake has quit IRC | 01:23 | |
| *** jj0hns0n has quit IRC | 01:26 | |
| n88 | 1) is there anybody still here and 2) has somebody got dashboard working? | 01:26 |
| *** jj0hns0n has joined #openstack | 01:26 | |
| haji | anybody here knows how to configure a cisco switch with openstack | 01:27 |
| vidd | 1.) yes 2.)no | 01:27 |
| btorch | hsom: what version of swift are you using ? this auth-server.conf error kind of reminds me of really old swift release when devauth was still used and there was an auth-server daemon if I recall correctly | 01:27 |
| btorch | hsom: nevermind skipped some of the conversation lines with notmyname | 01:28 |
| *** miclorb__ has quit IRC | 01:29 | |
| n88 | vidd: thanks :-P | 01:30 |
| vidd | n88, its not that i cant get horizon to work.... | 01:30 |
| n88 | vidd: I got it up and running and can log in but I've got something mis-configured with keystone or dashboard b/c I'm getting I don't have permission errors | 01:30 |
| vidd | keystone is my issue | 01:30 |
| n88 | vidd: maybe we are in the same boat…can you see any actual information in dashboard? | 01:30 |
| vidd | wow...what a coincidence | 01:30 |
| vidd | n88, let me ask you a different question.... | 01:31 |
| n88 | vidd: shoot! | 01:31 |
| vidd | did you test euca before installing keystone? | 01:31 |
| n88 | vidd: test euca? you mean eucalyptus cloud stuff? or euca commands? | 01:32 |
| vidd | euca commands | 01:32 |
| n88 | vidd: I found **most** of them worked | 01:32 |
| vidd | i use euca-describe-availability-zones verbose --debug | 01:32 |
| vidd | that is my "tester" | 01:32 |
| vidd | before keystone...all smiles | 01:32 |
| vidd | after keystone, they wont even talk to me | 01:33 |
| n88 | vidd: let me guess…after keystone…no smiles? | 01:33 |
| vidd | not even no smiley | 01:33 |
| n88 | vidd: all XXX or not even a listing? | 01:33 |
| vidd | if i got all XXX it would be a step in the right direction | 01:33 |
| livemoon | morning vidd | 01:33 |
| vidd | hello livemoon | 01:34 |
| n88 | vidd: I actually just tried that command and realized I'm still using the —use_deprecated_auth in my nova.conf | 01:34 |
| vidd | n88 im getting a timeout | 01:34 |
| *** troya has joined #openstack | 01:34 | |
| n88 | vidd: if I comment that line out is there any more nova.conf flags you've set for keystone? | 01:34 |
| vidd | n88 did you get the smileys? | 01:34 |
| livemoon | yesterday I thought using private cloud to replace LAMP | 01:34 |
| n88 | vidd: I'm still using the old authentication b/c I haven't changed my nova.conf…but I'm about to do that | 01:35 |
| n88 | vidd: is there any other nova.conf settings youv'e chnaged for keystone? | 01:35 |
| vidd | n88, my question was not"do you have the deprec auth flag...the question was did you get the smileys | 01:35 |
| n88 | also…does anyone know where dashboard stores its log files? I know that seems like a silly question but I can't seem to find them | 01:35 |
| troya | hi alll | 01:35 |
| vidd | n88, they are in your apache error logs =] | 01:36 |
| vidd | its a web site after all | 01:36 |
| *** Turicas has joined #openstack | 01:36 | |
| troya | hi vidd and livemoon | 01:36 |
| vidd | hello troy | 01:36 |
| vidd | hello troya | 01:36 |
| n88 | vidd: I'm getting smiles right now after installing Keystone but that's b/c I believe I'm still using Secret and access key authentication b/c I haven't commented out --use_deprecated_auth | 01:36 |
| n88 | vidd: ah right…yeah… *smacks self in head* | 01:37 |
| vidd | n88 its not as obvious as one might think =] | 01:37 |
| *** coulda has joined #openstack | 01:37 | |
| troya | vidd, i get new server now, by the way what's the best OS for openstack ? | 01:37 |
| troya | Centos or UBuntu ? | 01:37 |
| coulda | hi,how to config Multiple nodes with a single adapter? | 01:37 |
| vidd | since Red Hat scares the b'Jesus out of me, and i used to work with the xubuntu dev team...im not sure im one to give an unbiased answer | 01:38 |
| n88 | vidd: lol | 01:38 |
| *** nyeates has joined #openstack | 01:39 | |
| vidd | coulda, you can [try] using my scripts as long as you opt-out of keystone and dashboard | 01:40 |
| vidd | my keystone and dashboard scripts are causing me issues i have to track down yet | 01:41 |
| vidd | coulda, https://github.com/vidd/vidd | 01:42 |
| *** v0id has joined #openstack | 01:42 | |
| *** dwcramer has quit IRC | 01:45 | |
| coulda | vidd,i have readed network_setup.sh ,bu i want to use flatnetwork | 01:47 |
| vidd | coulda, dont take this the wrong way...but why? why would you not want to have your private ip's automatically assigned to your instaces by your controller node? | 01:48 |
| coulda | vidd ,i am doing proof of concept in vmware+NAT | 01:49 |
| *** haji has quit IRC | 01:52 | |
| vidd | coulda, i would suggest at least attempting the dhcp route first | 01:52 |
| vidd | but in theory, you set your /etc/networking/interfaces up with iface eth0 manual | 01:54 |
| vidd | then set up your br100 with your eth0 info | 01:55 |
| vidd | *iface eth0 inet manual | 01:55 |
| *** jj0hns0n has quit IRC | 01:56 | |
| *** jj0hns0n has joined #openstack | 01:56 | |
| coulda | vidd ,i use flatnetwork,but I am getting the following error DataSourceEc2.py[WARNING]: ‘http://169.254.169.254′ failed: url error, | 01:56 |
| *** jj0hns0n has quit IRC | 01:57 | |
| coulda | vidd ,-A nova-network-PREROUTING -d 169.254.169.254/32 -p tcp -m tcp --dport 80 -j DNAT --to-destination xxxx:8773，But no effect | 01:57 |
| *** ohnoimdead has quit IRC | 01:57 | |
| hsom | Exception: Could not create account AUTH_system for user system:root | 01:57 |
| hsom | using tempauth ^ | 01:58 |
| vidd | coulda, your opening up port 80 for port 8773 traffic | 01:58 |
| *** dwcramer has joined #openstack | 01:58 | |
| *** troya has quit IRC | 01:58 | |
| vidd | thats not going to work | 01:59 |
| *** adjohn has quit IRC | 01:59 | |
| coulda | vidd ,i use tcpdump,but i do not capture any data | 02:00 |
| vidd | can you open all the ports? | 02:00 |
| coulda | vidd，yes | 02:01 |
| vidd | and your ip range looks weird .... why specify a /32 or is that the way IP tables have to be? | 02:03 |
| livemoon | who know how to put scripts into instance when it start | 02:04 |
| *** PeteDaGuru has quit IRC | 02:04 | |
| *** bcwaldon has quit IRC | 02:06 | |
| *** jog0 has joined #openstack | 02:07 | |
| *** brian_sc1 has joined #openstack | 02:07 | |
| *** anotherjesse has quit IRC | 02:07 | |
| *** brian_sc2 has joined #openstack | 02:08 | |
| *** brian_sc1 has quit IRC | 02:08 | |
| coulda | vidd,you can see config in http://paste.openstack.org/show/3383/ | 02:09 |
| *** brian_sc has quit IRC | 02:11 | |
| vidd | coulda, you have no eth ports specified in your nova.conf | 02:11 |
| *** dolphm has quit IRC | 02:11 | |
| vidd | look at my nova-preload https://github.com/vidd/vidd/blob/master/nova-preload.sh | 02:12 |
| n88 | vidd: Is your euca-describe-availability showing it trying to use the keystone service? Or do you have a Secret/Access key? | 02:12 |
| n88 | does euca-describe-instances even support keystone? I don't think it does | 02:13 |
| vidd | n88, im sure it does | 02:13 |
| vidd | n88, you getting smileys sill? | 02:13 |
| *** jdurgin has quit IRC | 02:16 | |
| coulda | vidd,which flag? | 02:17 |
| *** chadh has quit IRC | 02:17 | |
| *** chadh has joined #openstack | 02:17 | |
| vidd | coulda, the last 4 in the networking section | 02:17 |
| *** jakedahn has joined #openstack | 02:17 | |
| coulda | vidd,ok,i try it | 02:19 |
| *** brian_sc2 has quit IRC | 02:19 | |
| coulda | vidd,thanks | 02:19 |
| vidd | just pay the bill when ya get it in the mail =] | 02:19 |
| *** jog0 has left #openstack | 02:20 | |
| *** osier has joined #openstack | 02:20 | |
| *** jog0 has joined #openstack | 02:20 | |
| *** jog0 has quit IRC | 02:20 | |
| *** jog0 has joined #openstack | 02:20 | |
| n88 | vidd: I get smileys but I can't launch instances | 02:21 |
| vidd | n88 your 3 steps ahead of me =\ | 02:21 |
| *** andrewbogott has quit IRC | 02:22 | |
| *** coulda has quit IRC | 02:22 | |
| vidd | everything was fine until keystone shoed up =\ | 02:22 |
| *** CaptTofu has quit IRC | 02:22 | |
| vidd | *showed | 02:23 |
| n88 | vidd: yeah…well this whole openstack suite is definitely got some growing pains | 02:23 |
| vidd | i keeps this up its going to get some lead poisoning =] | 02:24 |
| vidd | *it keeps.... | 02:24 |
| *** gyee has quit IRC | 02:26 | |
| n88 | well this is odd…did you realize dashboard installs nova-compute? | 02:27 |
| n88 | well at least Kiall's PPAs do | 02:28 |
| vidd | n88, nova-compute? or novaclient? | 02:28 |
| n88 | nova-compute | 02:28 |
| vidd | n88, you should already HAVE nova-compute | 02:29 |
| n88 | I'll have to let him know…unless he has a good reason for his openstack-dashboard project to have a dependency on nova-compute | 02:29 |
| vidd | nova-compute is what lets you launch instances | 02:29 |
| vidd | i should say "let's you host VM's | 02:30 |
| n88 | vidd: yes, but I'm running a cloud controller and 3 compute nodes and my cloud controller isn't supposed to run nova-compute | 02:30 |
| *** deshantm_laptop_ has joined #openstack | 02:30 | |
| n88 | esp since it can't…it's a 7 year old machine with no VT-extensions :-PO | 02:31 |
| vidd | n88 it still needs nova-compute | 02:31 |
| *** reed has joined #openstack | 02:34 | |
| *** deshantm_laptop has quit IRC | 02:34 | |
| *** jdg has joined #openstack | 02:34 | |
| *** woleium has quit IRC | 02:36 | |
| *** miclorb_ has joined #openstack | 02:36 | |
| *** anotherjesse has joined #openstack | 02:41 | |
| *** anotherjesse has quit IRC | 02:42 | |
| *** jdg has quit IRC | 02:43 | |
| hugokuo | does anyone have idea about using Bittorrent to spread an object in Swift ? | 02:43 |
| *** jdg has joined #openstack | 02:44 | |
| *** Turicas has quit IRC | 02:52 | |
| *** stewart has quit IRC | 02:53 | |
| livemoon | hugokuo: you have so advanced | 02:54 |
| livemoon | vidd | 02:54 |
| hugokuo | ? | 02:54 |
| vidd | livemoon, ?? | 02:55 |
| livemoon | hugo:ÄãÒѽ›ºÜ³¬Ç°ÁË | 02:55 |
| livemoon | vidd: do you know how to put scripts into instance when instance is been started | 02:55 |
| vidd | livemoon, no | 02:56 |
| * vidd barely knows how to get scripts onto real computers =] | 02:56 | |
| vidd | livemoon, you may want to look at puppet | 02:57 |
| vidd | livemoon, yoiu have keystone working.... | 02:57 |
| livemoon | yes, and you? | 02:57 |
| vidd | no...i cant get it working now | 02:57 |
| livemoon | cool | 02:58 |
| n88 | livemoon: look at cloud-init if you're running ubuntu instances | 02:58 |
| vidd | i dont know what i did differently =\ | 02:58 |
| livemoon | thanks | 02:58 |
| livemoon | you are repo+ git? | 02:58 |
| n88 | livemoon: I know some people also couple cloud-init with puppet…I have only used cloud-init | 02:58 |
| vidd | livemoon, no...pure Kiall ppa's | 02:58 |
| livemoon | what about aother os? | 02:59 |
| *** supriya_ has joined #openstack | 02:59 | |
| livemoon | rhel, centos? | 02:59 |
| vidd | i have to be doing something wrong with my converting to keystone, because i can run euca commands before keystone but not after | 03:00 |
| *** clauden_ has quit IRC | 03:01 | |
| *** dwcramer has quit IRC | 03:06 | |
| *** clauden_ has joined #openstack | 03:06 | |
| *** jdg has quit IRC | 03:06 | |
| *** ejat has joined #openstack | 03:13 | |
| *** ejat has joined #openstack | 03:13 | |
| *** woleium has joined #openstack | 03:13 | |
| *** nati2 has quit IRC | 03:16 | |
| *** clopez has quit IRC | 03:18 | |
| *** dwcramer has joined #openstack | 03:18 | |
| *** ejat has quit IRC | 03:20 | |
| *** miclorb_ has quit IRC | 03:21 | |
| *** ejat has joined #openstack | 03:21 | |
| *** ejat has joined #openstack | 03:21 | |
| *** nati2 has joined #openstack | 03:22 | |
| *** Ryan_Lane has quit IRC | 03:28 | |
| *** cdub has quit IRC | 03:32 | |
| *** miclorb_ has joined #openstack | 03:34 | |
| *** nati2 has quit IRC | 03:36 | |
| *** adjohn has joined #openstack | 03:37 | |
| *** stewart has joined #openstack | 03:40 | |
| *** ldlework has quit IRC | 03:43 | |
| *** mdomsch has joined #openstack | 03:45 | |
| *** haji has joined #openstack | 03:47 | |
| vidd | livemoon, where did you get your keystone? | 03:50 |
| livemoon | git | 03:50 |
| vidd | in your databse, do you have a "service_id " field for "roles" | 03:51 |
| Kiall | vidd: still have ec2+keystone issues? | 03:52 |
| Kiall | http://paste.openstack.org/show/3384/ | 03:52 |
| Kiall | thats the client side covered... | 03:52 |
| Kiall | then, in keystone creds table key = $user:$project and secret = $password | 03:53 |
| vidd | Kiall, is the region needed? | 03:53 |
| Kiall | Yea - for some reason it seems to be | 03:53 |
| Kiall | But, its not used by euca commands | 03:54 |
| Kiall | and yes, i have service_id column, but all NULL's | 03:54 |
| *** vladimir3p has quit IRC | 03:55 | |
| Kiall | If matching those doesn't work, then make sure your actually using all my PPA packages.. pastebin `dpkg -l | grep -E (openstack|nova|glance|keystone)` and i can check... | 03:55 |
| vidd | so NOT $AWSkeys:$prject | 03:55 |
| Kiall | I use $username:$project as my $awskey | 03:56 |
| Kiall | and $password as my $awssecret | 03:56 |
| Kiall | but, they can be anything... | 03:56 |
| Kiall | pop in some random uuid's if you like etc | 03:56 |
| Kiall | give `dpkg -l | grep -E "(openstack|nova|glance|keystone)"` a go, EC2 issues are something fixed in stable/diablo and not the ubuntu packages.. | 03:59 |
| Kiall | you could be using one of those by mistake | 03:59 |
| vidd | passlib 1.5.2-1~openstack0~oneiric1 | 03:59 |
| Kiall | there should be a pile more than that ;) | 03:59 |
| Kiall | http://paste.ubuntu.com/740843/ | 03:59 |
| *** krow has quit IRC | 04:00 | |
| vidd | http://dpaste.com/657809/ | 04:01 |
| Kiall | no openstackx? | 04:01 |
| vidd | nope | 04:01 |
| vidd | Unpacking openstackx (from .../openstackx_0.2.0~20111021-0mit5_all.deb | 04:02 |
| Kiall | bounce nova-api after its done | 04:02 |
| vidd | bounced them all | 04:03 |
| Kiall | I wonder if I should just make that a nova-api dependency | 04:03 |
| *** deshantm_laptop_ has quit IRC | 04:03 | |
| *** deshantm_laptop has joined #openstack | 04:04 | |
| vidd | Kiall, yeah | 04:04 |
| *** sdake has joined #openstack | 04:04 | |
| *** miclorb_ has quit IRC | 04:04 | |
| Kiall | my nova api-paste http://paste.ubuntu.com/740848/ | 04:04 |
| * Kiall loves the "pastebinit" command ;) | 04:05 | |
| vidd | ??pastebinit?? | 04:06 |
| Kiall | yea .. `pastebinit /etc/nova/nova.conf` or .. `dpkg -l | grep -E "(openstack|nova|glance|keystone)" | pastebinit` | 04:07 |
| Kiall | it sends it up and pops a URL out | 04:07 |
| Kiall | anyway, any joy with openstackx installed? or my api-paste? | 04:09 |
| vidd | nope | 04:10 |
| Kiall | you used my exact api-paste, with just the service-token changed? | 04:10 |
| vidd | yep | 04:12 |
| vidd | no joy | 04:12 |
| vidd | can you run euca-describe-availability-zones verbose ? | 04:12 |
| Kiall | oh! | 04:12 |
| vidd | or is this just something NOT to use as a test =\ | 04:12 |
| Kiall | --keystone_ec2_url=http://78.153.XXX.XXX:5000/v2.0/ec2tokens in nova.conf? | 04:12 |
| vidd | yes | 04:13 |
| Kiall | yea - euca-describe-availability-zones verbose works | 04:13 |
| *** misheska has quit IRC | 04:13 | |
| vidd | Kiall, you want to look at my https://github.com/vidd/vidd/blob/master/convert_keystone.sh | 04:14 |
| vidd | before this peaches and cream...after dead | 04:14 |
| Kiall | my nova.conf, maybe there is something in it | 04:14 |
| Kiall | http://paste.ubuntu.com/740852/ | 04:14 |
| Kiall | you called the repo "vidd" lol ;) | 04:15 |
| *** robbiew has quit IRC | 04:15 | |
| vidd | it did it all on its own =\ | 04:16 |
| Kiall | lies | 04:16 |
| vidd | ok...so i didit accidentally cuzz i have no idea how this daggone github uploader doodad worked | 04:17 |
| Kiall | I don't see anything off in that script.. | 04:17 |
| vidd | [and im still not that sure] | 04:17 |
| Kiall | what do you have in the keystone credentials table? | 04:17 |
| vidd | i did have $EC2_ACCESS_KEY:$proj and $EC2_SECRET_KEY | 04:19 |
| Kiall | you should have: user_id = 1, tenant_id = 1, type = EC2, key = admin:admin, secret = password | 04:19 |
| Kiall | assuming user_id=admin | 04:19 |
| Kiall | tenant_id=admin | 04:19 |
| *** wariola has joined #openstack | 04:19 | |
| *** nyeates has quit IRC | 04:19 | |
| Kiall | admin's password=password | 04:19 |
| Kiall | Thats assuming the ENV vars I pasted earlier using | 04:20 |
| vidd | then i changed it to $NOVA_USERNAME:$NOVA_PROJECT_ID and $NOVA_API_KEY | 04:20 |
| Kiall | so, what exactly is in your EC2 vars, and your keystone creds table.. (password blocked out ofc) | 04:20 |
| *** andrewbogott has joined #openstack | 04:22 | |
| *** nati2 has joined #openstack | 04:23 | |
| vidd | right now.... creds table is: id =1 user_id =1, tenant_id = 1 type=EC2 key=vidd:viddnet secret= vidd | 04:23 |
| Kiall | and EC2 env vars? | 04:24 |
| *** dolphm has joined #openstack | 04:24 | |
| vidd | EC2_SECRET_KEY=xxxxxx-xxxx-xxxx-xxxxxx-xxxxxx | 04:25 |
| Kiall | uhh | 04:25 |
| Kiall | sorry? | 04:25 |
| vidd | EC2_USER_ID=42 | 04:25 |
| Kiall | Stop. | 04:26 |
| Kiall | EC2_SECRET_KEY=vidd | 04:26 |
| Kiall | EC2_ACCESS_KEY=vidd:viddnet | 04:26 |
| Kiall | working? | 04:28 |
| hsom | anyone here familiar w/ swift? | 04:28 |
| hsom | i'm having a HELLUVA time gettin git running. i have some general questions | 04:29 |
| vidd | Kiall, nope | 04:29 |
| *** bcwaldon has joined #openstack | 04:29 | |
| vidd | still timing out | 04:29 |
| Kiall | timing out? | 04:29 |
| Kiall | sitting there hanging? | 04:29 |
| vidd | 2011-11-16 23:28:45,382 euca2ools [DEBUG]:received 500 response, retrying in 2 seconds | 04:30 |
| Kiall | run it with --verbose (or is it --debug) if its hanging | 04:30 |
| vidd | i always do that | 04:30 |
| vidd | thats how i know its timing out | 04:30 |
| Kiall | was this a fresh ubuntu install when you started with the packages? | 04:30 |
| vidd | yep | 04:30 |
| * vidd is not afraid to format a machine [6x a day][ | 04:31 | |
| Kiall | pastebin /usr/share/pyshared/keystone/middleware/ec2_token.py will ya? ;) | 04:31 |
| vidd | http://paste.ubuntu.com/740862/ | 04:32 |
| vidd | wow...that pastebinit IS cool =] | 04:32 |
| Kiall | ;) | 04:32 |
| Kiall | Okay - that has the fix in it ;) | 04:32 |
| vidd | i should hope so...its your stuff =] | 04:33 |
| Kiall | can you scrap any EC2 env vars you have bar EC2_SECRET_KEY, EC2_ACCESS_KEY and EC2_URL | 04:33 |
| Kiall | and, what is your EC2_URL set to? | 04:33 |
| Kiall | those are the only 3 EC2 vars I have set, and it works for me ;) | 04:34 |
| Kiall | Maybe one of the others like EC2_USER_ID=42 is doing something weird | 04:34 |
| vidd | EC2_URL=http://192.168.15.200:8773/services/Cloud | 04:34 |
| Kiall | yea - thats right.. | 04:34 |
| vidd | i dont know how to deep-6 an env =\ | 04:35 |
| Kiall | env --unset EC2_USER_ID etc | 04:35 |
| vidd | =\ wtf | 04:36 |
| Kiall | it works? | 04:36 |
| vidd | thats not working | 04:36 |
| vidd | the env is staying | 04:36 |
| Kiall | oh .. maybe that command is wrong then¬ | 04:36 |
| *** miclorb_ has joined #openstack | 04:36 | |
| Kiall | vidd: sorry its "unset EC2_bla" | 04:37 |
| hsom | swift anyone? | 04:39 |
| Kiall | hsom: sorry, never used it | 04:39 |
| hsom | :( | 04:39 |
| v0id | hsom, answers.launchpad.net/swift | 04:40 |
| *** Ryan_Lane has joined #openstack | 04:41 | |
| Kiall | vidd: unsetting them help? | 04:42 |
| vidd | Kiall, this is the garbage i get http://paste.ubuntu.com/740868/ | 04:42 |
| Kiall | hows the time sync between everything BTW? | 04:42 |
| vidd | Kiall, maybe i should have you run my scripts in a vm and see if they work for ya =] | 04:43 |
| Kiall | lol .. | 04:43 |
| vidd | its all on one machine | 04:43 |
| Kiall | Or you run mine, they work :P | 04:43 |
| vidd | i only have one nic' | 04:43 |
| Kiall | vidd: pastebin your nova.conf | 04:43 |
| Kiall | *maybe* there is something ... | 04:44 |
| vidd | i doubt it | 04:44 |
| *** scottjg has joined #openstack | 04:44 | |
| Kiall | try me ;) | 04:45 |
| vidd | http://dpaste.com/657819/ | 04:45 |
| Kiall | uhh, I hope dpaste just messed that up? | 04:46 |
| Kiall | --keystone_ec2_url=192.168.15.200:5000/v2.0/ec2tokens | 04:46 |
| Kiall | --ajax_console_proxy_url=http:192.168.15.200//:8000 | 04:46 |
| *** cdub has joined #openstack | 04:47 | |
| vidd | nope...those them | 04:47 |
| vidd | url needs to have http | 04:47 |
| Kiall | exactly. | 04:48 |
| Kiall | Your one's dont | 04:48 |
| *** dolphm_ has joined #openstack | 04:48 | |
| Kiall | and your console proxy url has slashes in the middle of it ;) | 04:48 |
| *** bcwaldon has quit IRC | 04:49 | |
| vidd | alright! we have progress | 04:50 |
| vidd | 2011-11-16 23:49:44,508 euca2ools [ERROR]:401 Unauthorized | 04:50 |
| Kiall | OMG | 04:50 |
| Kiall | That was the issue -_- | 04:50 |
| *** dolphm has quit IRC | 04:50 | |
| Kiall | while (respect > 0) respect--; | 04:51 |
| vidd | ? | 04:51 |
| Kiall | Anyway, you have EC2_SECRET_KEY=vidd and EC2_ACCESS_KEY=vidd:viddnet now? | 04:53 |
| vidd | Kiall, yes...in both keystone and env | 04:54 |
| vidd | and now i have 401 error | 04:54 |
| vidd | at least its not timing out anymore | 04:55 |
| Kiall | and what role have you given the user? | 04:55 |
| *** hadrian has quit IRC | 04:55 | |
| vidd | all roles | 04:55 |
| vidd | admin, keystoneadmin and member | 04:55 |
| vidd | admin with a Capital A | 04:56 |
| Kiall | i have: Admin w/ tenant NULL, Admin w/ tenant 1, KeystoneAdmin w/ tenant NULL, KeystoneServiceAdmin w/ tenant NULL | 04:56 |
| vidd | my user has all 3 roles | 04:57 |
| Kiall | Yes, but 4 entries in the user_roles table | 04:57 |
| Kiall | (Not sure if they are all needed, but whatever) | 04:57 |
| vidd | i only have 3 roles =\ | 04:58 |
| *** hsom has quit IRC | 04:59 | |
| vidd | Admin, KeystoneServiceAdmin, Member | 04:59 |
| Kiall | well 4 roles, of which the user has 2 of them with tenant = NULL and 1 with both tenant = NULL and tenant = 1 | 04:59 |
| Kiall | Thats just what I have, frankly, I don't know if they are all needed | 04:59 |
| *** dolphm has joined #openstack | 05:00 | |
| Kiall | I stole the devstack sample data.. | 05:00 |
| vidd | Admin is needed do euca stuff, member is needed to doe horizon stuff, and keystoneServiceAdmin is not active yet | 05:00 |
| Kiall | Well, The devstack sample data works for me.. ;) | 05:02 |
| Kiall | https://github.com/managedit/openstack-setup/blob/master/keystone_data.sh.tmpl | 05:02 |
| Kiall | (they devstack one has changed recently) | 05:02 |
| *** dolphm_ has quit IRC | 05:03 | |
| livemoon | hi | 05:03 |
| livemoon | i am back | 05:03 |
| vidd | is --FAKE_subdomain=ec2 needed? | 05:05 |
| Kiall | nope. | 05:05 |
| Kiall | my nova.conf - http://paste.ubuntu.com/740852/ | 05:05 |
| *** dwcramer has quit IRC | 05:08 | |
| vidd | Kiall, i fixed my typo's you want to take my scripts for a spin in a VM? | 05:08 |
| *** vladimir3p has joined #openstack | 05:09 | |
| Kiall | I'd much prefer you try the devstack sample data before I go to all that effort ;) | 05:09 |
| *** nati2 has quit IRC | 05:12 | |
| *** miclorb_ has quit IRC | 05:13 | |
| *** dolphm has quit IRC | 05:15 | |
| haji | can you tell me anything about cloudaudit? when can i use it with openstack | 05:15 |
| *** sandywalsh_ has quit IRC | 05:16 | |
| *** Ryan_Lane has quit IRC | 05:17 | |
| *** dolphm_ has joined #openstack | 05:25 | |
| *** CaptTofu has joined #openstack | 05:26 | |
| vidd | Kiall, so, there is some thing in that data that works | 05:27 |
| Kiall | ? | 05:27 |
| Kiall | As in, you have it working? | 05:27 |
| vidd | i have to figure out what it is because i cant use garbage data | 05:27 |
| Kiall | delete the user_roles entries 1 at a go from the DB, restarting KS after each go.. | 05:28 |
| Kiall | might tell you ;) | 05:28 |
| vidd | or...just add this KeystoneAdmin role | 05:30 |
| Kiall | ;) | 05:30 |
| vidd | you think it would be nice of the ppl that put these apps together to be so kind as to pre-load required roles when you first fire up the app =\ | 05:31 |
| Kiall | I'm pretty sure they accept patches ;) | 05:31 |
| vidd | i'll take a combat ribbon =\ | 05:32 |
| vidd | oh...and the garbage data didnt work out of the box either | 05:33 |
| Kiall | actually, you can barely use git.. you probably shouldn't bother trying to submit a patch with gerrit ;) | 05:33 |
| Kiall | does for me :P | 05:33 |
| *** Satya has joined #openstack | 05:33 | |
| vidd | i had to completely dump the entire database an load it fresh | 05:34 |
| vidd | and manually add the endpoints | 05:34 |
| vidd | and the passwords were messed up | 05:34 |
| vidd | but....other than that! | 05:35 |
| vidd | worked like a charm | 05:35 |
| Kiall | well, at least it works ;) | 05:36 |
| vidd | so...my issue is not with the convert...its with the template load | 05:36 |
| vidd | no...it doesnt work....it functions | 05:37 |
| Kiall | ..and the convert mucks up nova.conf ;) | 05:37 |
| vidd | no...the nova-preload mucked up the nova config | 05:37 |
| *** chomping has quit IRC | 05:38 | |
| vidd | the convert just turned the broken piece on | 05:38 |
| vidd | and apparently, the ajax thing wasnt even in the mix =] | 05:38 |
| Kiall | ;) | 05:39 |
| vidd | so...i think my issue with my keystone_template is just the missing KeystoneAdmin role | 05:46 |
| vidd | i think everything else is golden | 05:46 |
| vidd | oh...and i have to pull the prompt for a password | 05:46 |
| *** v0id has quit IRC | 05:53 | |
| *** rnorwood has joined #openstack | 05:53 | |
| *** haji has quit IRC | 05:54 | |
| *** rsampaio has quit IRC | 05:58 | |
| *** deshantm_laptop has quit IRC | 06:00 | |
| *** brian_sc has joined #openstack | 06:02 | |
| *** rnorwood has quit IRC | 06:03 | |
| *** brian_sc has left #openstack | 06:03 | |
| *** nerens has joined #openstack | 06:07 | |
| *** vladimir3p has quit IRC | 06:07 | |
| *** sophiap has quit IRC | 06:09 | |
| *** dolphm_ has quit IRC | 06:28 | |
| *** dolphm has joined #openstack | 06:30 | |
| *** supriya_ has quit IRC | 06:32 | |
| *** kaigan has joined #openstack | 06:33 | |
| vidd | Kiall, you still here? | 06:34 |
| *** andrewbogott has quit IRC | 06:35 | |
| *** objectiveous has joined #openstack | 06:36 | |
| *** objectiveous has left #openstack | 06:36 | |
| livemoon | hi | 06:37 |
| vidd | hello livemoon | 06:38 |
| livemoon | vidd | 06:38 |
| livemoon | I remove 70-persistent-net.rules, but after reboot it is created | 06:39 |
| vidd | ok....i have no idea what you mean | 06:40 |
| livemoon | oh | 06:40 |
| *** guigui1 has joined #openstack | 06:43 | |
| *** CaptTofu has quit IRC | 06:44 | |
| *** wariola has quit IRC | 06:45 | |
| *** shaon has joined #openstack | 06:49 | |
| *** jedi4ever has joined #openstack | 06:52 | |
| *** wariola has joined #openstack | 06:53 | |
| *** Bryanstein has quit IRC | 07:00 | |
| *** jog0 has quit IRC | 07:02 | |
| *** mnour has quit IRC | 07:02 | |
| *** mindpixel has joined #openstack | 07:07 | |
| *** dolphm has quit IRC | 07:08 | |
| *** wariola has quit IRC | 07:09 | |
| *** obino has joined #openstack | 07:09 | |
| *** woleium has quit IRC | 07:14 | |
| *** mgoldmann has joined #openstack | 07:14 | |
| *** wariola has joined #openstack | 07:21 | |
| *** shaon has quit IRC | 07:22 | |
| *** wariola has quit IRC | 07:23 | |
| *** wariola has joined #openstack | 07:27 | |
| *** matt_ has joined #openstack | 07:28 | |
| *** matt_ is now known as Guest57027 | 07:28 | |
| *** dolphm has joined #openstack | 07:28 | |
| vidd | dolphm, hello | 07:29 |
| vidd | dolphm, you had expressed an interest in my scripts some hours ago...i was wondering if you had an opertunity to look at them | 07:31 |
| *** dolphm has quit IRC | 07:32 | |
| *** wariola has quit IRC | 07:33 | |
| *** Guest57027 has left #openstack | 07:34 | |
| *** lucrecius has joined #openstack | 07:36 | |
| *** jksd has joined #openstack | 07:38 | |
| *** lucrecius has left #openstack | 07:39 | |
| *** Bryanstein has joined #openstack | 07:40 | |
| *** vidd is now known as vidd-away | 07:45 | |
| *** wariola has joined #openstack | 07:46 | |
| *** jksd has left #openstack | 07:47 | |
| *** adjohn has quit IRC | 07:55 | |
| *** wariola has quit IRC | 07:57 | |
| *** shaibn has joined #openstack | 07:58 | |
| *** mrevell has joined #openstack | 08:01 | |
| *** lucrecius has joined #openstack | 08:03 | |
| livemoon | hi | 08:04 |
| *** MarkAtwood has quit IRC | 08:04 | |
| *** vladimir3p has joined #openstack | 08:04 | |
| *** derjohn_mob has joined #openstack | 08:07 | |
| *** wariola has joined #openstack | 08:07 | |
| *** vladimir3p has quit IRC | 08:09 | |
| *** lucrecius has quit IRC | 08:09 | |
| *** sophiap has joined #openstack | 08:10 | |
| *** jog0 has joined #openstack | 08:18 | |
| *** MarkAtwood has joined #openstack | 08:19 | |
| *** Satya has quit IRC | 08:20 | |
| *** mrevell has quit IRC | 08:21 | |
| *** TheOsprey has joined #openstack | 08:25 | |
| *** vernhart has joined #openstack | 08:27 | |
| *** vernhart1 has quit IRC | 08:30 | |
| *** foexle has joined #openstack | 08:31 | |
| foexle | hiho | 08:31 |
| *** jog0 has quit IRC | 08:33 | |
| *** superjudge_ has joined #openstack | 08:33 | |
| *** shaon has joined #openstack | 08:37 | |
| *** katkee has quit IRC | 08:38 | |
| *** redconnection has joined #openstack | 08:42 | |
| *** nacx has joined #openstack | 08:42 | |
| *** alekibango_ has quit IRC | 08:43 | |
| *** alekibango_ has joined #openstack | 08:43 | |
| *** chemikadze has quit IRC | 08:46 | |
| *** chemikadze has joined #openstack | 08:48 | |
| *** pixelbeat has joined #openstack | 08:52 | |
| *** Razique has joined #openstack | 08:54 | |
| *** PotHix has joined #openstack | 08:56 | |
| *** koolhead17 has joined #openstack | 08:56 | |
| *** koolhead17 has joined #openstack | 08:57 | |
| *** superjudge_ has quit IRC | 08:58 | |
| *** superjudge_ has joined #openstack | 08:58 | |
| *** TreeStump has joined #openstack | 09:06 | |
| *** javiF has joined #openstack | 09:12 | |
| *** epic_k has quit IRC | 09:15 | |
| *** dobber has joined #openstack | 09:19 | |
| *** javiF has quit IRC | 09:19 | |
| *** javiF has joined #openstack | 09:19 | |
| *** cdub has quit IRC | 09:21 | |
| *** foexle has quit IRC | 09:22 | |
| *** foexle has joined #openstack | 09:22 | |
| *** darraghb has joined #openstack | 09:23 | |
| koolhead17 | hi all | 09:23 |
| livemoon | hi,koolhead17 | 09:24 |
| koolhead17 | hi livemoon | 09:24 |
| livemoon | have you think use openstack as replace traditional datacenter arch | 09:25 |
| livemoon | for example. replace LAMP in one machine | 09:29 |
| *** shaon has quit IRC | 09:31 | |
| *** clopez has joined #openstack | 09:33 | |
| *** scottjg has quit IRC | 09:34 | |
| *** reed has quit IRC | 09:37 | |
| Razique | hi all | 09:39 |
| livemoon | hi | 09:39 |
| livemoon | Razique | 09:39 |
| *** cdub has joined #openstack | 09:40 | |
| *** clopez has quit IRC | 09:40 | |
| *** saju_m has joined #openstack | 09:41 | |
| *** flamboz has quit IRC | 09:44 | |
| *** ejat has quit IRC | 09:47 | |
| koolhead17 | hey Razique | 09:49 |
| *** smulcahy has joined #openstack | 09:49 | |
| smulcahy | morning | 09:49 |
| smulcahy | is there any pointers to a minimum spec for a small openstack test cluster? | 09:49 |
| guaqua_ | what kind of stuff do you wont to do? | 09:50 |
| smulcahy | would 1 x cloud controller, 1 x storage and 3 x compute node be reasonable? | 09:50 |
| guaqua_ | just nova or also swift? | 09:50 |
| guaqua_ | nova needs 2 nodes | 09:50 |
| smulcahy | quaqua_: basically r&d to figure out if the stuff we're running on ec2 atm could be moved onto a private cloud | 09:50 |
| guaqua_ | nova = compute, swift = storage | 09:51 |
| *** guaqua_ is now known as guaqua | 09:51 | |
| smulcahy | so need something that could be migrated to semi-production at some stage | 09:51 |
| smulcahy | I guess I want to run both nova and swift so? | 09:51 |
| guaqua | you probably want 3 separate nodes for tsorage | 09:51 |
| smulcahy | do I need swift for a small private cloud? | 09:51 |
| guaqua | and run proxy on some or all of them | 09:51 |
| smulcahy | or it is only if I want s3 type functionality? | 09:51 |
| guaqua | yep | 09:51 |
| foexle | hi Razique | 09:52 |
| guaqua | no need if you don't store lots and lots of data | 09:52 |
| smulcahy | ah ok | 09:52 |
| smulcahy | still, I guess we need to play around with it anyway | 09:52 |
| guaqua | if you move in the under-terabyte -range | 09:52 |
| guaqua | or just don't need scaling just yet | 09:52 |
| smulcahy | we do terabytes well here | 09:52 |
| Razique | hey foexle | 09:52 |
| guaqua | hmm | 09:52 |
| Razique | koolhead17: livemoon what's up ? | 09:52 |
| smulcahy | so maybe worth spec'ing some storage anyway | 09:52 |
| *** katkee has joined #openstack | 09:53 | |
| smulcahy | so 1 x cc, 2 x nova and 1 x swift would give a working minimal config? | 09:53 |
| guaqua | in general, you want 3 replicas, which means 3 zones, which means at least 3 nodes for storage | 09:53 |
| *** reed has joined #openstack | 09:53 | |
| smulcahy | ah ok | 09:53 |
| koolhead17 | Razique, fighting with proxy to enable pip install :P | 09:53 |
| guaqua | you can run swift on a single node as well | 09:53 |
| guaqua | you just need to scale it up later, set it up again | 09:53 |
| smulcahy | right | 09:53 |
| smulcahy | yeah, thats do-able and maybe better initial use of hw | 09:54 |
| guaqua | haven't tried myself, if you can actually up the number of replicas | 09:54 |
| guaqua | if you do just 1 replica and 1 storage node running everything | 09:54 |
| guaqua | and then incrementally add more to the cluster and rebalance it | 09:54 |
| guaqua | you can add more nodes to a running cluster in general, but changing the replica count is sort of 'undefined' but what i understood is that it's just not a common use case | 09:55 |
| guaqua | running 1 storage nodes just doesn't give you redundancy, but you can test all the interfaces and all that | 09:55 |
| guaqua | do the integration part | 09:55 |
| guaqua | and then dive into swift specifics later on | 09:56 |
| guaqua | hopefully someone else provides their opinion aswell :) | 09:56 |
| guaqua | i'm not running a production setup, yet | 09:56 |
| ogelbukh | i'm sorry, but couldn't we have 3 replicas on one server? | 09:56 |
| smulcahy | thats a good start anyway, thanks a lot for the feedback | 09:56 |
| guaqua | ogelbukh: well, yes. taking 3 times the disk space :) | 09:56 |
| ogelbukh | ^) | 09:57 |
| uvirtbot | ogelbukh: Error: ")" is not a valid command. | 09:57 |
| livemoon | koolhead17 , Razique: have you used swift? | 09:57 |
| ogelbukh | but easier to scale later | 09:57 |
| guaqua | you can run many "nodes" on a single host | 09:57 |
| Razique | livemoon: nope, never | 09:57 |
| guaqua | yeah, easier to scale later | 09:57 |
| ogelbukh | ok | 09:57 |
| ogelbukh | thanks ) | 09:57 |
| ogelbukh | just wanted to clarify | 09:57 |
| guaqua | yeah | 09:57 |
| smulcahy | in terms of machine spec | 09:58 |
| smulcahy | what are people running here typically? | 09:58 |
| smulcahy | dual quad core, 32GB ram type machines? | 09:58 |
| smulcahy | or bigger/smaller? | 09:58 |
| ogelbukh | i'm working on a pretty same configuration now ) one node for swift to be scaled later | 09:58 |
| guaqua | you should look at what rackspace runs | 09:58 |
| guaqua | ogelbukh: you have multiple storage nodes on one physical node? | 09:58 |
| ogelbukh | i'm evaluating this possibility right now ) | 09:59 |
| guaqua | okay :) | 09:59 |
| smulcahy | http://www.referencearchitecture.org/hardware-specifications/ | 09:59 |
| smulcahy | hhmm | 09:59 |
| guaqua | good to hear others doing this stuff | 09:59 |
| guaqua | i'm planning a 4 node start for a swift cluster | 09:59 |
| *** troya has joined #openstack | 09:59 | |
| smulcahy | not sure the budget will stretch for nodes with 96GB of ram just now | 09:59 |
| smulcahy | :) | 09:59 |
| guaqua | ram is mostly for memcached if i understand correctly | 10:00 |
| guaqua | and disc cache | 10:00 |
| reidrac | guaqua: 4U storage nodes with 24 disks are very popular :) | 10:00 |
| guaqua | i'm thinking of going for something a lot less radical | 10:00 |
| reidrac | 4 storage nodes, and a couple of proxies it's a good start for a production cluster | 10:01 |
| guaqua | starting small and replacing them when we have actual billing | 10:01 |
| reidrac | there's a problem with that, because you can't change the number of partitions in the ring | 10:01 |
| troya | hi all | 10:01 |
| guaqua | reidrac: i.e. can't have different kind of nodes in the ring or? | 10:02 |
| reidrac | you can start with a large number of partitions because you expect to add drives later | 10:02 |
| troya | hi reidrac | 10:03 |
| reidrac | guaqua: yes you can, but the number of partitions can be changed once the ring is created | 10:03 |
| guaqua | can or can't be? | 10:03 |
| guaqua | isn't this just a planning decision, not as much a technical one? | 10:03 |
| reidrac | you can have different kind of nodes | 10:03 |
| guaqua | i.e. plan for x amount of storage? | 10:04 |
| reidrac | the main planning issue with swift is that you need to calculate how many partitions you want to have in advance | 10:04 |
| reidrac | and that value is important because if p is partitions and s is cluster size, p is constant | 10:05 |
| reidrac | so if you increase s, that means the partitions will be bigger... but not that you'll have more partitions | 10:05 |
| reidrac | ans that affects directly to the replication | 10:05 |
| reidrac | *and | 10:05 |
| livemoon | hi, troy | 10:05 |
| guaqua | slower replication? | 10:05 |
| livemoon | see you once in the morning and see another time in the evening | 10:06 |
| reidrac | if you partitions are too big, that would mean that in the event of losing a data (ie. three disks broken in three different zones) | 10:06 |
| reidrac | you'll lose more content | 10:06 |
| guaqua | uhh. i need to clear this stuff up for myself. the concept of a partition isn't quite clear enough | 10:06 |
| reidrac | and yes... I guess replication performance can be affected | 10:06 |
| *** MarkAtwood has left #openstack | 10:07 | |
| reidrac | guaqua: I agree. It's not a disk partition but a logical partition inside the cluster | 10:07 |
| guaqua | hmm. so, basically, you'd like to have partition size optimized so that they are not too small and not too large | 10:08 |
| reidrac | yep | 10:08 |
| reidrac | based in the maximum size of your cluster | 10:08 |
| guaqua | having them too small means overhead, having them too large makes the probability of losing a partition due to a disk failure higher | 10:08 |
| reidrac | let's say you're using 4U 24 disks storage nodes with 2TB disks | 10:08 |
| *** Satya has joined #openstack | 10:09 | |
| reidrac | you can do the maths based on the probability of disk failure, the number of storage nodes, the number of replicas | 10:09 |
| Satya | Hi all | 10:09 |
| reidrac | and guess an average file size | 10:09 |
| Satya | i have some issue | 10:09 |
| reidrac | etc | 10:09 |
| reidrac | :D | 10:09 |
| guaqua | reidrac: okay | 10:10 |
| guaqua | reidrac: this is good stuff | 10:10 |
| Satya | if a instance is rebuilded from a image which is in saving state its goes to rebuild state after if it rebuilded with an active state good image it fails to come to active state | 10:11 |
| reidrac | guaqua: swift is pretty cool when something goes wrong, but getting the numbers right it's definitely very important | 10:11 |
| guaqua | reidrac: okay | 10:11 |
| Satya | and stays in rebuild state always | 10:11 |
| guaqua | reidrac: what kind of cluster are you running? | 10:11 |
| guaqua | and for what use? | 10:11 |
| reidrac | we have different clusters at the moment, but we when with 4 zones and 3 replicas based on our numbers | 10:12 |
| Satya | any help on it !! is it a possible bug? | 10:12 |
| reidrac | ie. the expected probability of disk failure | 10:13 |
| *** donatello92 has joined #openstack | 10:13 | |
| smulcahy | thanks all for feedback | 10:13 |
| *** smulcahy has quit IRC | 10:13 | |
| reidrac | s/when/went/ | 10:13 |
| reidrac | guaqua: docs recommends a hw spec from rackspace experience, it's worth looking at it! | 10:14 |
| *** reed has quit IRC | 10:14 | |
| ogelbukh | reidrac: thanks for hints! | 10:14 |
| *** miclorb_ has joined #openstack | 10:16 | |
| guaqua | reidrac: yeah, i'll go over it again :) | 10:16 |
| guaqua | reidrac: and thanks for this! very good stuff | 10:16 |
| reidrac | you're welcome | 10:16 |
| *** clopez has joined #openstack | 10:18 | |
| *** nerens has quit IRC | 10:19 | |
| *** dirkx_ has joined #openstack | 10:21 | |
| *** nerens has joined #openstack | 10:23 | |
| *** donatello92 has left #openstack | 10:24 | |
| *** dirkx_ has quit IRC | 10:24 | |
| *** dirkx_ has joined #openstack | 10:25 | |
| *** dpippenger has quit IRC | 10:25 | |
| *** Frodent has joined #openstack | 10:26 | |
| *** FroMaster has quit IRC | 10:26 | |
| *** aliguori has quit IRC | 10:26 | |
| Razique | anyone have succeed in instances suspend ? | 10:27 |
| *** aliguori has joined #openstack | 10:27 | |
| *** redconnection has quit IRC | 10:27 | |
| *** dpippenger has joined #openstack | 10:28 | |
| *** troya has quit IRC | 10:28 | |
| koolhead17 | Razique, i think i did, when i used from source | 10:29 |
| Razique | the novaclient or nova-compute ? | 10:29 |
| koolhead17 | nova-compute | 10:30 |
| Razique | good to know | 10:32 |
| *** dirkx_ has quit IRC | 10:32 | |
| *** superjudge_ has joined #openstack | 10:33 | |
| *** Satya has quit IRC | 10:37 | |
| *** donatello92 has joined #openstack | 10:42 | |
| *** donatello92 has left #openstack | 10:43 | |
| *** Satya has joined #openstack | 10:45 | |
| koolhead17 | Razique, did horizon added few deps as well? | 10:45 |
| koolhead17 | new | 10:45 |
| Razique | deps ? | 10:46 |
| livemoon | hi | 10:46 |
| Satya | hi | 10:46 |
| livemoon | who knows git and puppet well? | 10:46 |
| *** livemoon has left #openstack | 10:49 | |
| *** GheRivero has joined #openstack | 10:50 | |
| *** wariola has quit IRC | 10:51 | |
| *** ahasenack has joined #openstack | 11:04 | |
| *** vernhart1 has joined #openstack | 11:09 | |
| Razique | ok guys i found out the suspend issue | 11:09 |
| Razique | I needed to update etc/apparmor.d/usr.lib.libvirt.virt-aa-helper | 11:10 |
| Razique | and add this http://paste.openstack.org/show/3385/ | 11:10 |
| *** vernhart has quit IRC | 11:12 | |
| *** livemoon has joined #openstack | 11:13 | |
| livemoon | hi | 11:13 |
| *** viraptor has joined #openstack | 11:14 | |
| *** adjohn has joined #openstack | 11:15 | |
| viraptor | how does the 'osapi_extension' flag work? I see there's a "magic" extension "standard_extensions" which loads everything else, but if I wanted to load only some specific ones, what would I have to put there? list of extension classes, or modules? | 11:15 |
| *** livemoon has quit IRC | 11:27 | |
| *** adjohn has quit IRC | 11:37 | |
| *** vernhart1 has quit IRC | 11:38 | |
| *** FloFri has joined #openstack | 11:41 | |
| FloFri | hello everyone, its me again ^^ . have still the problem with the metadata injection | 11:42 |
| FloFri | Razique: are you there? | 11:42 |
| *** cdub has quit IRC | 11:48 | |
| *** dirkx_ has joined #openstack | 11:51 | |
| *** livemoon has joined #openstack | 11:57 | |
| *** FloFri has quit IRC | 11:57 | |
| *** miclorb_ has quit IRC | 12:01 | |
| * koolhead17 getting dumber as day progressing :( | 12:02 | |
| koolhead17 | when am using keystone for dashboard, i have to do a syncdb and provide user credentials how are they handled in kestone? | 12:03 |
| koolhead17 | **keystone | 12:03 |
| livemoon | ？ | 12:05 |
| foexle | koolhead17: credentials | 12:06 |
| foexle | table | 12:06 |
| koolhead17 | foexle, of keystone ? | 12:06 |
| foexle | yep | 12:06 |
| foexle | its for assotiation of keystone and nova | 12:07 |
| foexle | -of + for | 12:07 |
| *** PeteDaGuru has joined #openstack | 12:07 | |
| koolhead17 | foexle, but when am installing dashboard without keystone i had to run syncdb for credentials. | 12:07 |
| *** vdo has joined #openstack | 12:08 | |
| foexle | hmmm i think dashboard will not run wihtout keystone auth | 12:08 |
| foexle | because you need to login (Username/PW) dashboard send this creds to keystone and keystone look in the credentials table which access and private key this user has in nova | 12:09 |
| koolhead17 | my question is the file local_settings.py | 12:10 |
| koolhead17 | OPENSTACK_KEYSTONE_URL = "http://localhost:5000/v2.0/" | 12:11 |
| koolhead17 | 12:11 | |
| foexle | i dont know if is possible to sync the novadb creds to keystone with a command | 12:11 |
| koolhead17 | OPENSTACK_KEYSTONE_ADMIN_URL = "http://localhost:35357/v2.0" | 12:11 |
| foexle | ahhhh :D | 12:11 |
| koolhead17 | will point to my keystone server | 12:11 |
| foexle | yes thats right | 12:11 |
| foexle | dont forget the last slash ! | 12:11 |
| *** katkee has quit IRC | 12:18 | |
| *** katkee has joined #openstack | 12:18 | |
| *** praefect has joined #openstack | 12:19 | |
| *** vdo has quit IRC | 12:23 | |
| *** tryggvil_ has quit IRC | 12:31 | |
| *** stewart has quit IRC | 12:32 | |
| Razique | back | 12:39 |
| *** reed has joined #openstack | 12:41 | |
| foexle | Razique: wb | 12:42 |
| pvo | foexle: there was a script to take the novadb creds and put them into keystone but it probably hasn't been touched in a while and keystone has gone through some changes. | 12:42 |
| foexle | pvo: oh yes ? where i can find this script ? | 12:43 |
| pvo | it was a nova-manage option. looking it up right now. | 12:43 |
| foexle | ah ok .... i'll find it thanks | 12:44 |
| pvo | ./bin/nova-manage shell export | 12:44 |
| pvo | but it looks broken | 12:44 |
| pvo | so, no help there. sorry | 12:44 |
| *** marrusl_ has joined #openstack | 12:45 | |
| foexle | hmmm | 12:45 |
| foexle | ok | 12:45 |
| pvo | it may just need some tweaking to fix it though | 12:45 |
| *** marrusl_ has quit IRC | 12:46 | |
| *** Oneiroi has joined #openstack | 12:46 | |
| *** katkee has quit IRC | 12:48 | |
| livemoon | hi | 12:50 |
| *** tryggvil_ has joined #openstack | 12:50 | |
| Oneiroi | ok finally for 2 x 6.1 boxes built have nova installing now :) | 12:50 |
| Oneiroi | only taken a couple months >< | 12:50 |
| livemoon | has anyone use openstack with puppet? | 12:50 |
| pvo | livemoon: yea a few people do. | 12:51 |
| livemoon | pvo: use puppet deploy instance? | 12:51 |
| livemoon | or deploy node? | 12:51 |
| pvo | to deploy nodes that run openstack. | 12:51 |
| livemoon | I want to deploy instances | 12:51 |
| pvo | I don't use it to boot an instance | 12:51 |
| livemoon | if I have a linux image, I send "user data" to it when it is be started, and some puppte command in 'user data', | 12:52 |
| *** osier has quit IRC | 12:52 | |
| livemoon | then instance can do something,ie: install LAMP | 12:53 |
| pvo | something like that could work | 12:54 |
| livemoon | I think it can but I havenot do it | 12:54 |
| *** vidd-away is now known as vidd | 12:54 | |
| vidd | Kiall, ae you here? | 12:55 |
| livemoon | vidd: you are get up? | 12:55 |
| pvo | something may have to call the script after you get it booted | 12:55 |
| pvo | you can pass it in, but you'll have to find a place that it will exe on boot | 12:55 |
| pvo | for your distro | 12:55 |
| pvo | or ssh in and exe it after it is pinging | 12:55 |
| vidd | yeah ... time to get the kid off to school | 12:55 |
| livemoon | in nova, there is 'user_data' | 12:56 |
| livemoon | I think it can work | 12:56 |
| livemoon | morning vidd | 12:56 |
| vidd | who knew one child [age 6] could sound like a herd of elephants? | 12:56 |
| *** dwcramer has joined #openstack | 12:56 | |
| livemoon | vidd: ah | 12:56 |
| koolhead17 | am i doing sumthing wrong http://paste.ubuntu.com/741131/ | 12:57 |
| livemoon | koolhead17: I also have this case | 12:57 |
| koolhead17 | livemoon, what is the solution? | 12:58 |
| livemoon | no, I just reinstall | 12:58 |
| livemoon | have you use "sudo python setup.py install" | 12:58 |
| livemoon | please try to use "sudo python setup.py develop" replace "install" | 12:59 |
| *** katkee has joined #openstack | 13:00 | |
| *** dprince has joined #openstack | 13:03 | |
| koolhead17 | livemoon, ok. let me see it again :) | 13:06 |
| vidd | koolhead17, what i do is first do python setup install in the django-openstack folder | 13:07 |
| *** chemikadze has quit IRC | 13:07 | |
| koolhead17 | ok | 13:08 |
| vidd | then do the python setup.py install in the openstack-dashboard folder | 13:08 |
| vidd | and i never use the with_venv.sh | 13:09 |
| *** Razique_ has joined #openstack | 13:09 | |
| * vidd thinks keeping somethig with "magic reaches" into your stuff should not be "protected" from automatic security updates | 13:10 | |
| koolhead17 | and yes am uing lucid 10.04 | 13:10 |
| *** zparta has quit IRC | 13:10 | |
| koolhead17 | :) | 13:10 |
| vidd | koolhead17, should not matter....in theory, you can use the git meathod on windows | 13:11 |
| koolhead17 | indeed | 13:11 |
| *** pradeep has joined #openstack | 13:12 | |
| *** Razique has quit IRC | 13:12 | |
| *** Razique_ is now known as Razique | 13:12 | |
| koolhead17 | i remember last time i had to only sudo sh run_tests.sh | 13:13 |
| koolhead17 | and everything worked for me | 13:13 |
| *** yogi_ has joined #openstack | 13:15 | |
| *** zparta has joined #openstack | 13:16 | |
| *** kbringard has joined #openstack | 13:18 | |
| *** dwcramer has quit IRC | 13:20 | |
| *** mdomsch has quit IRC | 13:21 | |
| vidd | ok so i thin i have an issue with glance now =\ | 13:22 |
| livemoon | what? | 13:22 |
| vidd | i cant get anything to upload to it | 13:23 |
| livemoon | what error? | 13:24 |
| Oneiroi | nova.exception.ClassNotFound: Class Client could not be found: No module named dec :-/ *joy* | 13:25 |
| vidd | error: [Errno 111] ECONNREFUSED | 13:25 |
| livemoon | glance-api | 13:26 |
| vidd | full text here:http://dpaste.com/658079/ | 13:27 |
| livemoon | check glance-api process | 13:27 |
| livemoon | ps aux | grep glance | 13:27 |
| livemoon | check your keystone process | 13:27 |
| vidd | all good there | 13:28 |
| *** daniel__ has joined #openstack | 13:30 | |
| livemoon | netstat -lntp | 13:31 |
| livemoon | all port is open? | 13:31 |
| vidd | http://dpaste.com/658080/ | 13:32 |
| daniel__ | hello .... could anyone give me a hint if it is possible to activate the ec2 api in an openstack installation with keystone ... is that possible ... and if yes, where do i have to configure that? | 13:32 |
| vidd | daniel__, yes...it is possible | 13:32 |
| vidd | everythig work elsewhere? | 13:33 |
| livemoon | vidd: have you change anything in glance? | 13:33 |
| livemoon | and can it upload well before? | 13:33 |
| vidd | livemoon, i set the flags to keystone...but other than that no | 13:33 |
| vidd | no..never tried to upload | 13:34 |
| livemoon | set your glance-registry.conf and glance-api.conf | 13:34 |
| vidd | daniel__, make sure your user is set to admin role with NO tenant set | 13:34 |
| livemoon | no set , paste | 13:34 |
| daniel__ | yeah ... everything else works .... nova-api works fine ... dshboard and glance as well .... i also have set ec2 credentials in keystone ... but ehn i start keystone ... the only api's that are started are the nova service and nova admin api ... not the ec2 | 13:35 |
| foexle | vidd: really ? no tenant ? if i look in the keystone db i see a tenant is set in the user table | 13:35 |
| foexle | vidd: if i set to NULL i cant login anymore via dashboard | 13:36 |
| vidd | and add a credential as such "credentials add $NOVA_USERNAME EC2 $EC2_ACCESS_KEY $EC2_SECRET_KEY $NOVA_PROJECT_ID | 13:36 |
| *** dwcramer has joined #openstack | 13:36 | |
| vidd | foexle, your user needs to be in the "member" role to use dashboard | 13:37 |
| koolhead17 | vidd, you used mysql or sqlite for as database 4 dashboard | 13:37 |
| koolhead17 | ? | 13:37 |
| vidd | koolhead17, mysql | 13:37 |
| *** pradeep has quit IRC | 13:37 | |
| koolhead17 | vidd, and you install python-mysql from ubuntu pkg ? | 13:37 |
| livemoon | I am ready to sleep | 13:38 |
| koolhead17 | am having same issue again | 13:38 |
| foexle | vidd: yeah he is ... but without a tenant id, i don't get a login | 13:38 |
| koolhead17 | livemoon, see ya | 13:38 |
| daniel__ | user is set to admin with no tenant | 13:38 |
| foexle | livemoon: bye | 13:38 |
| vidd | foexle, its the "add user roles" section that you skip the tenant....not the add user" section | 13:39 |
| livemoon | vidd: it seems your glance never connect to keystone | 13:39 |
| livemoon | vidd: try to use glance -A token index | 13:39 |
| livemoon | to see if it is the same error | 13:39 |
| daniel__ | vidd, is there any tutorial how to start keystone with ec2? | 13:39 |
| vidd | livemoon, yep | 13:40 |
| vidd | daniel__, no | 13:40 |
| vidd | i just told you how | 13:40 |
| livemoon | i think your glance config need to modify | 13:40 |
| daniel__ | i had an installation via devstack .... users roles are good i think and ec2 credentials are set via credentials add | 13:41 |
| vidd | daniel__, do your EC2 creds match your ENV?? | 13:41 |
| *** deshantm_laptop has joined #openstack | 13:41 | |
| daniel__ | vidd, i think so ... but i dont even know where the ec2 endpoint is ... i diddnt set it in any configfile ... and after i start keystone there seems to be nothing at port 8773 | 13:43 |
| vidd | daniel__, run this keystone-manage credentials add $NOVA_USERNAME EC2 $EC2_ACCESS_KEY $EC2_SECRET_KEY $NOVA_PROJECT_ID | 13:43 |
| daniel__ | okay ... ill try that ... thanks for the moment | 13:44 |
| vidd | no substitutions...just copy paste as is | 13:44 |
| *** dwcramer has quit IRC | 13:44 | |
| vidd | it will pull the info from yoiur ENV so they match | 13:44 |
| daniel__ | from the .../keystone/bin dir? | 13:44 |
| vidd | and [in theory] you will be able to use EC2 commands | 13:45 |
| vidd | daniel__, if keystone-manage is not already in your path...yes | 13:45 |
| daniel__ | hmm ... ERROR: Unable to locate any configuration file. Cannot load application admin | 13:46 |
| vidd | daniel__, is you keystone even up? | 13:47 |
| daniel__ | yes .. dashboard works nova api woks | 13:48 |
| *** alekibango_ has quit IRC | 13:48 | |
| daniel__ | works | 13:48 |
| *** alekibango_ has joined #openstack | 13:48 | |
| foexle | can i use swift (cli) with keystone ? .... the keystone doc say yes, but i get a url parsing error | 13:50 |
| vidd | daniel__, what command did you enter? | 13:50 |
| daniel__ | localadmin@openstackmaster:/opt/stack/keystone/bin$ keystone-manage credentials add $NOVA_USERNAME EC2 $EC2_ACCESS_KEY $EC2_SECRET_KEY $NOVA_PROJECT_ID | 13:50 |
| vidd | daniel__, whay os? | 13:52 |
| daniel__ | ubuntu server 11.04 | 13:52 |
| daniel__ | openstack installation via devstack | 13:53 |
| *** ldlework has joined #openstack | 13:53 | |
| vidd | drop the /opt/stack/keytone/bin$ crap | 13:54 |
| vidd | keystone is in your path | 13:54 |
| vidd | you can run this command from anywhere on your system | 13:54 |
| daniel__ | hm okay ... | 13:56 |
| daniel__ | ... but that doesnt change the behaviour .... just get that error http://pastebin.com/PevncuDN | 13:58 |
| daniel__ | and that is the output after starting keystone ... http://pastebin.com/2P3N2FWh | 14:00 |
| *** dwcramer has joined #openstack | 14:00 | |
| vidd | daniel__, copy your keystone config file (wherever you currently have it) to /etc/keystone (you may need to mkdir that folder first) | 14:00 |
| n88 | vidd: hello…did you ever get your keystone working? | 14:01 |
| *** hadrian has joined #openstack | 14:02 | |
| daniel__ | keystone is installed in /opt/stack and not systemwide ....but okay ... ill try that ;) | 14:02 |
| *** redconnection has joined #openstack | 14:02 | |
| livemoon | you can set you own $DEST rather than /opt/stack | 14:02 |
| vidd | n88, yep...turns out that when you assign your user the "Admin" role, he CANNOT have a tenant specified | 14:02 |
| livemoon | vidd | 14:03 |
| daniel__ | SUCCESS: Credentials admin created. | 14:03 |
| vidd | daniel__, it IS installed systemwide | 14:03 |
| daniel__ | Yeah! Sounds good ... | 14:03 |
| livemoon | do you knonw 'user_data' in nova? | 14:03 |
| vidd | livemoon, not sure what you mean | 14:04 |
| *** ldlework has quit IRC | 14:04 | |
| vidd | daniel__, it just didnt know where to find its config file | 14:04 |
| livemoon | when you see dashboard launch instance | 14:05 |
| livemoon | there is a field name 'user data' | 14:05 |
| livemoon | It seems can input data to the instance | 14:05 |
| daniel__ | vidd, yes .. thats correct thanks! ... and now the ec2 endpoind is http://<ip>:8773/services/Cloud ?? | 14:05 |
| vidd | by default it looks "/etc/keystone/ then /etc , then ~/keystone, then ~/ | 14:05 |
| praefect | livemoon: I use user_data to configure instances on the fly | 14:05 |
| praefect | livemoon: I pass the user data through hybridfox, and the the VM reads this user_data and acts accordingly (self-config) | 14:06 |
| vidd | daniel__, no...no endpoint needed | 14:06 |
| livemoon | praefect: cool | 14:06 |
| daniel__ | how does the eucatools or any other ec2 client knows how to communicate? | 14:06 |
| livemoon | "user data" just set command in it? | 14:07 |
| praefect | livemoon: you can also pass binary files in user_data, https://help.ubuntu.com/community/CloudInit I use it with ubuntu and also windows instances | 14:07 |
| vidd | daniel__, EC2 already knows how to to the stuff...it just needs the nod from keystone | 14:07 |
| livemoon | praefect: what about centos or rhel? | 14:08 |
| vidd | dashboard has no clue how to do euca stuff...so it needs an endpoint from keystone | 14:08 |
| praefect | livemoon: check the link I sent you, it's basically a BLOB, you can pass whatever you need and receive it on the other side, there is a strict limit (1MB I think) | 14:08 |
| livemoon | praefect: thanks | 14:08 |
| daniel__ | vidd ... no not for dashboard ... ill try to run a ec2 api to use eucatools with openstack | 14:09 |
| * koolhead17 kicks himself | 14:09 | |
| praefect | livemoon: basically, any OS you have, you just need to query http://169.254.169.254/latest/metadata/user_data (or something like that), and the cloud will return the user-data you passed to this particular instance, you then do whatever you want | 14:09 |
| vidd | daniel__, you dont understand..... | 14:09 |
| praefect | livemoon: it's just well documented and well organised in ubuntu but obviously you can do it with any system | 14:10 |
| livemoon | :) | 14:10 |
| daniel__ | vidd, kinda ;) sry ... | 14:10 |
| daniel__ | so how could i connect a ec2 client to keystone? | 14:10 |
| vidd | daniel__, endpoints are for application A to talk to application B | 14:11 |
| vidd | daniel__, you already did | 14:11 |
| vidd | your done | 14:11 |
| vidd | it should work | 14:11 |
| *** msivanes has joined #openstack | 14:11 | |
| *** ldlework has joined #openstack | 14:11 | |
| *** msivanes has left #openstack | 14:12 | |
| livemoon | bye,all' | 14:12 |
| daniel__ | what is with EC2_URL in my ENV | 14:12 |
| livemoon | see you tomorrow | 14:12 |
| praefect | livemoon: ciao | 14:12 |
| vidd | ? | 14:13 |
| vidd | daniel__, what do you mean? | 14:13 |
| daniel__ | i cant understand how to configure the eucatools to connect to keystone .... | 14:13 |
| *** livemoon has left #openstack | 14:13 | |
| daniel__ | or for exaple hybridfox .... | 14:13 |
| daniel__ | example | 14:13 |
| vidd | daniel__, does dashboard work? | 14:14 |
| daniel__ | yes | 14:14 |
| daniel__ | worked all the time | 14:14 |
| vidd | so that means euca is ALREADY configured to work with keystone | 14:14 |
| vidd | acually...i should say keystone knows how to work with euca =] | 14:15 |
| daniel__ | so could i use a ec2 client now to use the cloud? | 14:15 |
| daniel__ | for example hybridfox from another host .... | 14:16 |
| vidd | daniel__, yes...in theory | 14:16 |
| vidd | you test by doing it from CLI | 14:16 |
| daniel__ | and every ec2 client needs an EC2_URL | 14:17 |
| daniel__ | or am i wrong? | 14:17 |
| *** rods has joined #openstack | 14:17 | |
| vidd | the url you need you can find [assuming your local CLI works] by typing env |grep EC2_URL=http | 14:18 |
| daniel__ | that is the url i set in my .bashrc | 14:19 |
| daniel__ | for example ... that command: | 14:19 |
| *** msivanes has joined #openstack | 14:19 | |
| daniel__ | euca-describe-availability-zones | 14:19 |
| daniel__ | diddnt work | 14:19 |
| vidd | daniel__, thats because the command is euca-describe-availability-zones verbose | 14:20 |
| daniel__ | no the command diddnt even end .. | 14:21 |
| *** foexle has quit IRC | 14:21 | |
| daniel__ | no output ... and i have to ctrl+c it | 14:21 |
| daniel__ | after ctrl+c that error occurs http://pastebin.com/DNHmutrW | 14:22 |
| *** ejat has joined #openstack | 14:22 | |
| *** ejat has joined #openstack | 14:22 | |
| vidd | then you either do not have a user $NOVA_USERNAME in your keystone database, or he is not set to the "Admin" role or he IS set to the "Admin" role and his tenant was included | 14:23 |
| daniel__ | okay | 14:24 |
| koolhead17 | yeah | 14:24 |
| vidd | koolhead17, so how you doing? | 14:25 |
| koolhead17 | vidd, same error | 14:25 |
| daniel__ | in the keystone database ..... there is a user admin .... the following way: | 14:26 |
| vidd | daniel__, | 14:26 |
| vidd | i did not ask about your admoin user | 14:26 |
| vidd | i asked about the user we gave credentials to | 14:26 |
| vidd | daniel__, do you have a "Role" = "Admin" | 14:27 |
| daniel__ | yes .. with id 1 | 14:28 |
| vidd | Capital A not lower case a | 14:28 |
| vidd | because "Admin" != "admin" | 14:29 |
| *** andrewbogott has joined #openstack | 14:30 | |
| vidd | do you have a user that is the same as your $NOVA_USERNAME from your "env |grep NOVA | 14:30 |
| *** deshantm_laptop_ has joined #openstack | 14:30 | |
| daniel__ | oh ... thats it! I have Admin | 14:30 |
| daniel__ | no ... Admin is what i have to have, right? | 14:31 |
| vidd | daniel__, dont get ahead of me | 14:31 |
| *** deshantm_laptop_ has quit IRC | 14:32 | |
| *** deshantm_laptop_ has joined #openstack | 14:32 | |
| vidd | lets look at "roles" first | 14:32 |
| vidd | is there a role "Admin" | 14:33 |
| *** deshantm_laptop has quit IRC | 14:33 | |
| *** deshantm_laptop_ has quit IRC | 14:34 | |
| daniel__ | http://pastebin.com/3qyKa5zx that are the database tables | 14:34 |
| *** deshantm_laptop has joined #openstack | 14:34 | |
| vidd | does $NOVA_USERNAME = admin or demo? | 14:35 |
| *** dwcramer has quit IRC | 14:36 | |
| daniel__ | that are my exports http://pastebin.com/5QgU9cjc | 14:36 |
| daniel__ | so direct to your question ... $NOVA_USERNAME = admin | 14:37 |
| n88 | vidd: in your nova.conf files, you DO NOT have --use_deprecated_auth Correct? | 14:39 |
| vidd | n88 correct...but Kiall does | 14:39 |
| *** jaypipes has joined #openstack | 14:39 | |
| msivanes | ping bdill | 14:40 |
| *** hadrian has quit IRC | 14:40 | |
| n88 | vidd: does Kiall have his nova.conf posted somewhere? i wanted to suggest he put it up on git since it would be pretty useful to see | 14:41 |
| daniel__ | maybe i have to restart nova? now that keystone knows where its config is? | 14:41 |
| vidd | daniel__, do you have in user_roles table "user_id =1 role_id=1 tenant_id-NULL? | 14:41 |
| daniel__ | i just saw that there is a flag in nova.conf: --ec2_url=http://172.22.161.101:8773/services/Cloud | 14:42 |
| daniel__ | and I DO NOT have --use_deprecated_auth | 14:42 |
| n88 | vidd: so I know this may be repeating yourself…but what were all the modifications you had to make to Kiall's scripts/setup to get things working? | 14:42 |
| daniel__ | no i have 1, 1, 1, 1 | 14:43 |
| vidd | daniel__, there is your issue | 14:43 |
| daniel__ | could i just change that in the database? | 14:43 |
| vidd | tenant MUST = NULL for the Admin role | 14:43 |
| vidd | yes | 14:43 |
| daniel__ | or by keystone-manage? | 14:43 |
| n88 | vidd: you mean under the token table? | 14:44 |
| vidd | n88, check out my scrips on gethub | 14:44 |
| n88 | ok sure…what's your linky? | 14:44 |
| n88 | vidd: also…what exactly are 'tenants'? I understand users and roles…but tenants…is that like 'projects'? | 14:44 |
| vidd | n88 https://github.com/vidd/vidd | 14:45 |
| vidd | n88, be warned...my glance currently does NOT work | 14:45 |
| vidd | daniel__, do it right from the database | 14:45 |
| n88 | vidd: it's ok….I feel like it's the pushme-pullme from Dr. DoLittle (if that reference makes sense) | 14:46 |
| n88 | you get one thing working, something else may break | 14:46 |
| n88 | but I'll get it | 14:46 |
| *** rsampaio has joined #openstack | 14:47 | |
| vidd | n88 and if you catch something PLEASE let me know! | 14:47 |
| *** bryguy has quit IRC | 14:47 | |
| *** lorin1 has joined #openstack | 14:47 | |
| n88 | vidd: also what's the api-paste-nova? What's that file? Do I need it? Where does it go? | 14:48 |
| vidd | n88 my scripts take it, customize it for you and put it into your /etc/nova/api-paste.ini | 14:49 |
| vidd | n88, dont forget to read the README! | 14:51 |
| *** hadrian has joined #openstack | 14:51 | |
| *** bryguy has joined #openstack | 14:51 | |
| *** Oneiroi has quit IRC | 14:53 | |
| *** Oneiroi has joined #openstack | 14:53 | |
| *** Oneiroi has quit IRC | 14:53 | |
| *** Oneiroi has joined #openstack | 14:53 | |
| praefect | I'm setting up keystone... when I had endpointTemplates, do I have to enter %HOST_IP% or the actual IP of my server? | 14:53 |
| *** Oneiroi has quit IRC | 14:53 | |
| praefect | anyone has a good walkthrough for keystone setup (besides identity starter guide) | 14:54 |
| vidd | %HOST_IP% should be replaced.... | 14:54 |
| vidd | %TENANT_ID% should NOT be replaced | 14:54 |
| n88 | vidd: ok thanks…that was another question/issue I had | 14:54 |
| n88 | vidd: about %TENANT_ID% | 14:55 |
| praefect | wow... so easy! =) thanks vidd | 14:55 |
| *** dwcramer has joined #openstack | 14:56 | |
| *** robbiew has joined #openstack | 14:56 | |
| praefect | when setting up keystone endponts, is the region the nova "zone" or it's domething different entirely than my compute zone? | 14:57 |
| *** wilmoore has joined #openstack | 14:58 | |
| *** pradeep has joined #openstack | 14:58 | |
| vidd | i would set it as the nova zone | 14:58 |
| *** CaptTofu has joined #openstack | 14:59 | |
| *** AlanClark has joined #openstack | 14:59 | |
| vidd | praefect, that may be the answer to why my glance wont work =] | 14:59 |
| *** lts has joined #openstack | 15:00 | |
| *** CaptTofu has quit IRC | 15:00 | |
| praefect | vidd: yeah I'm thinking this region thing is something entirely different than the compute zone, after all, keystone encompasses more than just nova... | 15:01 |
| *** mattray has joined #openstack | 15:02 | |
| praefect | vidd: did you create services with "keystone-manage add service add nova compute ..." it is not even in the identity starter guide but it is in some other doc... | 15:02 |
| vidd | praefect, yes i did | 15:03 |
| *** wilmoore_ has joined #openstack | 15:03 | |
| vidd | i even wrote a script for it =] | 15:03 |
| *** wilmoore has quit IRC | 15:03 | |
| vidd | i wrote a script for everything | 15:03 |
| *** Shentonfreude has joined #openstack | 15:04 | |
| praefect | vidd: I'd really like to have a look at your script, even if you only managed to have keystone and nova linked, that's what I'm trying to do, could you paste it to paste.openstack.org? | 15:04 |
| praefect | vidd: that would be of great healp | 15:04 |
| vidd | take me from a machine with internet access to full [albiet not-fully function] stack | 15:04 |
| praefect | help | 15:04 |
| vidd | praefect, https://github.com/vidd/vidd | 15:05 |
| praefect | vidd: nice, I'm getting, I'm using puppet and I can build a cloud in no time.. and it always incorporated my latests recipes, works great for me | 15:05 |
| *** reed has quit IRC | 15:05 | |
| *** cmasseraf has joined #openstack | 15:05 | |
| praefect | vidd: thanks for the link | 15:05 |
| vidd | praefect, if you see a mistake PLEASE let me know! | 15:06 |
| cmasseraf | Hello everyone, I'm looking for a while now... where's the documentation abou installing openstack with one ethernet only? | 15:06 |
| vidd | praefect, just to let you know, my scripts pull from Kiall 's ppa | 15:06 |
| *** mdomsch has joined #openstack | 15:07 | |
| praefect | vidd: I won't be able to run your script as I'm on RHEL using packages but it will be of great help for sure... thanks again | 15:07 |
| *** rsampaio has quit IRC | 15:07 | |
| vidd | cmasseraf, such documentation does not exist so far as I have found | 15:07 |
| vidd | cmasseraf, it can be done...its just not recommeded or documented | 15:09 |
| cmasseraf | I've read about it in the Ubuntu site... but I could not find it.. I've 6 Servers where I'm going to install openstack however I only have one ethernet | 15:09 |
| n88 | vidd: a tenant in keystone = a project in nova, correct? | 15:09 |
| vidd | each server only has one ethernet adaptor? | 15:09 |
| *** mcclurmc has quit IRC | 15:09 | |
| cmasseraf | yes | 15:09 |
| vidd | n88 no, a tenant in keystone is a tenant in keystone | 15:10 |
| Razique | hey vidd n88 :) | 15:10 |
| *** mcclurmc has joined #openstack | 15:10 | |
| vidd | n88 in nova, can a user have more than one project? | 15:11 |
| vidd | can a project have more than one user? | 15:11 |
| vidd | [i do not know these answers] | 15:11 |
| Razique | vidd: yah a user can have multiple projects :) | 15:12 |
| vidd | but i can say with authority....a tenant in keystone != a project in nova | 15:12 |
| Razique | when u create a project, u define an admin | 15:12 |
| *** lionel has quit IRC | 15:12 | |
| cmasseraf | anyone know how to install openStack with only 1 thernet adapter? | 15:13 |
| *** whaley has joined #openstack | 15:13 | |
| vidd | Razique, so user A can have 15 projects and different creds for each project | 15:13 |
| *** lionel has joined #openstack | 15:13 | |
| vidd | cmasseraf, yes...i have done it | 15:13 |
| Razique | vidd: in fact you can have the same creds | 15:13 |
| vidd | glance and dashboard a re a little flacky though | 15:13 |
| Razique | since they appli per user | 15:13 |
| Razique | in fact what changes is the project/ region | 15:14 |
| *** GheRivero_ has joined #openstack | 15:14 | |
| Razique | eg nova --project_name=my_project boot --flavor --image server | 15:14 |
| *** heiko has joined #openstack | 15:14 | |
| vidd | Razique, so region in keystone [should]= project in nova? | 15:15 |
| Razique | nope | 15:16 |
| Razique | tenant in KS = project in nova | 15:16 |
| Razique | region = group of services | 15:16 |
| Razique | with api as endpoint | 15:16 |
| n88 | Raz: morning | 15:16 |
| heiko | another question :D ... never ending story hahahah, so it's sufficent to create a user with creds in keystone if i use keystone for auth ? | 15:16 |
| n88 | or afternoon | 15:16 |
| Razique | n88: :) | 15:17 |
| vidd | heiko, yes | 15:17 |
| *** heiko is now known as foexle | 15:17 | |
| foexle | vidd: thx | 15:17 |
| vidd | this is assuming you intend to use something like dashboard to do your dirty work | 15:17 |
| n88 | raz: does that mean my nova projects (e.g. 'project1') should match my tenant names in keystone? | 15:17 |
| Razique | n88: yup | 15:18 |
| Razique | in fact a nova-manage export (don't remember) allows you to generate keystone-mange commands | 15:19 |
| n88 | Raz: oh really? ah crap…haha ok let me try that first before I add all these entires into my keystone DB | 15:19 |
| Razique | eg projet "project1" would generate a "keystone-manage tenant create "project1"" | 15:19 |
| *** PiotrSikora has quit IRC | 15:23 | |
| n88 | Raz: as a follow-up if my 'cloudAdmin' is given global Admin roles in keystone then do they need to be specifically added to each project/tenant? Or is global role sufficient to give them access to all projects/tenants? | 15:23 |
| foexle | Razique: is that right now to check swift auth ? curl -d '{"passwordCredentials" | 15:23 |
| foexle | and to port 5000 ? | 15:24 |
| n88 | and are roles hierarchical? Meaning if I'm a global role Admin I also am a Member as well? | 15:24 |
| foexle | because the swift cli are broken i think | 15:25 |
| *** PiotrSikora has joined #openstack | 15:25 | |
| foexle | and the documentation are verry old i guess | 15:25 |
| vidd | is this correct (someone with keystone + glance) --storage_availability_zone=nova | 15:26 |
| *** code_franco has joined #openstack | 15:26 | |
| *** dendro-afk is now known as dendrobates | 15:27 | |
| *** duckhand has joined #openstack | 15:28 | |
| *** swill has joined #openstack | 15:28 | |
| *** daniel__ has quit IRC | 15:32 | |
| swill | I am trying to get swift3 to work with swiftauth and I am currently getting a 500 error: Exception: Could not autocreate account 'AUTH_1'. Has anyone seen this before? | 15:34 |
| *** saju_m has quit IRC | 15:34 | |
| foexle | soooo ok i think i have a bug in swift or in the doc of keystone | 15:34 |
| swill | foexie which keystone auth method are you using? | 15:35 |
| foexle | self.keystone_url = urlparse(conf.get('keystone_url')) => fail | 15:35 |
| swill | foexle: swiftauth i am guessing? | 15:35 |
| *** jeh has joined #openstack | 15:35 | |
| foexle | use = egg:keystone#swiftauth | 15:35 |
| foexle | yeah | 15:36 |
| foexle | should i use tokenauth ? | 15:36 |
| swill | no, use swift auth | 15:36 |
| swill | one second | 15:36 |
| swill | foexle: check my post in this bug, it may shed some slight on the situation: https://bugs.launchpad.net/keystone/+bug/890415 | 15:37 |
| foexle | swill: thx i'll check | 15:37 |
| swill | your url should be http://keystone_url:admin_port/v2.0 | 15:37 |
| jeh | I need to add an additional iptable rule to nova-network-snat, but it gets removed when a new instance is created. Anyone know how I can add additional rules to this on diablo? | 15:38 |
| swill | foexle: the admin port is configured in your keystone.conf file. | 15:38 |
| foexle | yes i know | 15:38 |
| foexle | swill: but i dont have a attribite with keystone url | 15:38 |
| foexle | because in doc isn't use this attribute | 15:38 |
| foexle | +the | 15:38 |
| swill | foexle: in the proxy-server.conf file? | 15:40 |
| foexle | so the answer from yogesh in your bug report is using other attributes in the keystone filter section as the actual doc | 15:40 |
| foexle | yes | 15:40 |
| foexle | look at the post of yogesh and the readme "swift integration" step 6 | 15:41 |
| foexle | its verry different | 15:41 |
| *** rustam has joined #openstack | 15:41 | |
| swill | what yogesh posted is what you should use | 15:42 |
| foexle | so are the doc incorrect | 15:42 |
| swill | assuming you changed the default admin_port to 5000 in your keystone.conf file. | 15:42 |
| foexle | yes yes :) i know ;) | 15:43 |
| swill | foexle: link the doc you are looking at | 15:43 |
| foexle | https://github.com/openstack/keystone | 15:43 |
| foexle | readme "swift integration" | 15:43 |
| foexle | step 6 | 15:43 |
| *** abecc has joined #openstack | 15:43 | |
| *** dgags has joined #openstack | 15:43 | |
| *** vladimir3p has joined #openstack | 15:44 | |
| swill | so that readme is for using tokenauth | 15:44 |
| swill | you will notice at the very bottom: Note: Keystone currently allows any valid token to do anything with any account. | 15:44 |
| swill | the new way of doing it so you do not have that issue is with swiftauth. | 15:45 |
| foexle | ahhh ok | 15:45 |
| foexle | ok ok thats a statement :) | 15:45 |
| foexle | thx | 15:45 |
| swill | the bug i referenced is to update the docs so the docs use swiftauth | 15:46 |
| *** redconnection has quit IRC | 15:46 | |
| swill | but the swiftauth docs are a little unclear, which is what my comment is about in that bug. | 15:46 |
| *** robbiew has quit IRC | 15:47 | |
| *** robbiew has joined #openstack | 15:47 | |
| n88 | vidd: you there? If so, I see in your config you add endpoints for all your tenants…do you have to do this? | 15:49 |
| *** wilmoore_ has quit IRC | 15:49 | |
| *** tryggvil_ has quit IRC | 15:49 | |
| foexle | swill: thanks :) thats a good explanation :) | 15:49 |
| *** tryggvil_ has joined #openstack | 15:49 | |
| vidd | n88, for each user...dunno about each tenant | 15:50 |
| *** GheRivero_ has quit IRC | 15:50 | |
| swill | foexle: no worries. if you run into other issues, ping me, i have been doing the same thing as you for the past few days. :) | 15:50 |
| *** jakedahn has quit IRC | 15:50 | |
| foexle | swill: all right :) | 15:51 |
| swill | has anyone run into a 500 error "Exception: Could not autocreate account 'AUTH_1'" when trying to get swift3 to work with swiftauth? | 15:53 |
| chmouel | swill: you prob didn't create the account first with swauth-add-user | 15:56 |
| *** alekibango_ has quit IRC | 15:57 | |
| *** MarkAtwood has joined #openstack | 15:57 | |
| *** alekibango_ has joined #openstack | 15:57 | |
| chmouel | actually you using tempauth or swauth? | 15:57 |
| swill | chmouel: i am using swiftauth, not swauth. | 15:57 |
| chmouel | chmouel: keystone swift auth right? | 15:58 |
| foexle | chmouel: he's using keystone | 15:58 |
| foexle | with swift | 15:58 |
| swill | chmouel: i have swiftauth working, but when i put swift3 in front of it, it stops working | 15:58 |
| chmouel | okay cool, so if autocreation are not working it's probably have to be that the account server cannot write on the fs | 15:58 |
| chmouel | check the logs on the storage servers | 15:58 |
| vidd | im getting an error when trying to talk to glance: http://dpaste.com/658130/ | 15:58 |
| chmouel | ah | 15:58 |
| chmouel | oh sorry | 15:58 |
| chmouel | yeah i'm woring on that 8-) | 15:59 |
| swill | chmouel: i think it has something to do with the tokens that swift3 tries to create are not going through (or something like that). | 15:59 |
| chmouel | yeah it's not supported yet | 15:59 |
| chmouel | need to do token validation on s3 first | 15:59 |
| swill | chmouel: ahhh, ok... | 15:59 |
| chmouel | on s3 token validation on keystone I mean | 15:59 |
| chmouel | there is a patch floating which I was integrating | 15:59 |
| *** wilmoore has joined #openstack | 16:00 | |
| chmouel | but had to fix other bug in keystone middleware first before addming that | 16:00 |
| swill | chmouel: ok, iw ill look around for the pathc. | 16:00 |
| *** imsplitbit has joined #openstack | 16:00 | |
| *** guigui1 has quit IRC | 16:00 | |
| *** mdomsch has quit IRC | 16:00 | |
| swill | chmouel: ya for sure. i will look into it and see if i can verify the patch works. | 16:01 |
| *** mdomsch has joined #openstack | 16:01 | |
| swill | and if not, i will see what is needed to fix it. | 16:01 |
| chmouel | well basically it's s3 tokens validation on keystone | 16:02 |
| swill | chmouel: will it have to happen in swiftauth (i am assuming) | 16:03 |
| *** javiF has quit IRC | 16:03 | |
| chmouel | no on keystone | 16:03 |
| *** dendrobates is now known as dendro-afk | 16:03 | |
| chmouel | and swiftauth will use that to validate the token | 16:03 |
| *** dendro-afk is now known as dendrobates | 16:04 | |
| swill | right... that makes sense. | 16:04 |
| foexle | swill: how do you create a new user if they exists in keystone but not in swift ? | 16:04 |
| chmouel | foexle: it use the autocreate feature of swift | 16:05 |
| *** dragondm has joined #openstack | 16:05 | |
| chmouel | so baiscally mean trust everything the auth give to me | 16:05 |
| *** ahasenack has quit IRC | 16:06 | |
| *** dendrobates is now known as dendro-afk | 16:06 | |
| *** ahasenack has joined #openstack | 16:07 | |
| *** wilmoore has quit IRC | 16:08 | |
| jeh | for reference I fixed my iptables rules by using POSTROUTING, instead of nova-network-snat | 16:08 |
| swill | chmouel: do you have a link to the patch for keystone to do s3 token validation? | 16:08 |
| *** rwmjones has quit IRC | 16:08 | |
| swill | chmouel: i have my setup installing with chef, so i can play with it and blow it away if i break things more. :) | 16:08 |
| *** freeflying has joined #openstack | 16:10 | |
| vidd | anyone here have glance working with keystone? | 16:11 |
| *** mindpixel has quit IRC | 16:12 | |
| foexle | swill: can you use "swfit" command ? | 16:12 |
| swill | foexle: the swift command you mean? | 16:13 |
| *** rnorwood has joined #openstack | 16:13 | |
| foexle | jop | 16:13 |
| chmouel | swill: that's the patch i am looking to integrate http://www.debian.or.jp/~yosshy/openstack-diablo/ but that's not prob going to be in this form | 16:13 |
| *** freeflyi1g has quit IRC | 16:13 | |
| swill | foexle: yes, i use: swift -A http://127.0.0.1:5000/v1.0 -U username -K password stat -v | 16:13 |
| *** obino has quit IRC | 16:14 | |
| foexle | i get a issue then | 16:14 |
| foexle | curl works fine | 16:14 |
| swill | chmouel: ok. i will wait for the official patch then. | 16:14 |
| swill | foexle: notice the /v1.0 | 16:14 |
| swill | not 2.0 | 16:14 |
| foexle | yeah | 16:14 |
| swill | 5000 is the keystone service_port in this case | 16:15 |
| foexle | http://pastebin.com/p1PnXLwt | 16:15 |
| foexle | yes i know | 16:15 |
| swill | checking | 16:15 |
| foexle | thats all correct | 16:15 |
| swill | that means that you are using the wrong url and keystone is not returning something that is useful | 16:16 |
| swill | or your endpoints are not setup correctly | 16:16 |
| foexle | °° | 16:16 |
| foexle | hmmm | 16:16 |
| foexle | -.- | 16:17 |
| vidd | anyone here have glance working with keystone? | 16:17 |
| swill | pastebin what you used to setup your keystone template endpoints | 16:17 |
| *** cp16net has joined #openstack | 16:17 | |
| foexle | RegionOne | 11 | http://192.168.0.2:5000/v2.0 | http://192.168.0.2:5001/v2.0 | http://192.168.0.2:5001/v2.0 | 16:17 |
| vidd | swill, you have keystone and glance working...right? | 16:17 |
| *** maplebed has quit IRC | 16:17 | |
| swill | vidd: not glance, just keystone and swift. | 16:18 |
| foexle | and yes keystone are running on all interfaces | 16:18 |
| swill | vidd: what method are you using tokenauth? | 16:18 |
| swill | foexle: that 11, is that identity? | 16:18 |
| swill | foexle: and what do you have for swift? | 16:18 |
| *** reidrac has quit IRC | 16:19 | |
| vidd | authtoken | 16:19 |
| foexle | swift | 16:19 |
| foexle | | 1 | RegionOne | 10 | http://%HOST_IP%:8080/v1/AUTH_%tenant_id% | http://%HOST_IP%:8080/ | http://%HOST_IP%:8080/v1/AUTH_%tenant_id% | 16:19 |
| swill | vidd: i have a working version of swift with authtoken, but i have not worked with glance yet. | 16:19 |
| foexle | and yes | 16:19 |
| foexle | | 11 | Keystone | identity | Keystone Identity Service | | 16:19 |
| swill | foexle: i believe you have to associate a tentant with each of your users otherwise you get this error. | 16:20 |
| swill | if you have done that, then it is an enpoint problem | 16:20 |
| swill | just remembered that. | 16:20 |
| swill | i will send you what iuse to setup my env. | 16:21 |
| foexle | yes my user has a tenant id | 16:21 |
| *** cmasseraf has quit IRC | 16:21 | |
| swill | http://pastebin.com/geA3ZqUP | 16:22 |
| swill | in your keystone.conf file, you have two roles defined, probably Admin and KeystoneServiceAdmin. do you have both of them defined in your keystone env? | 16:23 |
| foexle | yes | 16:23 |
| *** Badri has joined #openstack | 16:24 | |
| *** mikeyp has joined #openstack | 16:24 | |
| *** hadrian has quit IRC | 16:24 | |
| *** redconnection has joined #openstack | 16:24 | |
| foexle | but today it's fine :) ... i'm tired and hungry :D .... so cu later or tomorrow ! bye guys | 16:24 |
| *** rwmjones has joined #openstack | 16:24 | |
| *** rnorwood has quit IRC | 16:25 | |
| Badri | Has anyone tried nova boot with --key_path? I dont think it works. However --key_name seems to work. | 16:25 |
| swill | foexle: cheers. :) | 16:25 |
| foexle | swill: thx 4 your help i'll try tomorrow ;) | 16:25 |
| swill | foexle: no worries. gl | 16:25 |
| *** foexle has quit IRC | 16:25 | |
| *** cp16net has quit IRC | 16:27 | |
| *** bcwaldon has joined #openstack | 16:28 | |
| *** jeh has quit IRC | 16:29 | |
| *** jeh has joined #openstack | 16:29 | |
| Badri | Folks:: Has anyone tried nova boot with --key_path? I dont think it works. However --key_name seems to work. | 16:31 |
| *** mikeyp has quit IRC | 16:31 | |
| *** PiotrSikora has quit IRC | 16:31 | |
| *** dobber has quit IRC | 16:32 | |
| *** derjohn_mob has quit IRC | 16:32 | |
| *** cdub has joined #openstack | 16:32 | |
| *** PiotrSikora_ has joined #openstack | 16:32 | |
| *** ahale- has quit IRC | 16:32 | |
| *** viraptor has quit IRC | 16:32 | |
| *** redconnection has quit IRC | 16:33 | |
| BuZZ-T | does anyone know what causes this bug: https://bugs.launchpad.net/horizon/+bug/888385 | 16:35 |
| BuZZ-T | it happends if i try to install horizon on Ubuntu 11.10 via git | 16:35 |
| *** cp16net has joined #openstack | 16:36 | |
| BuZZ-T | "InstallationError: No files/directores in /home/rneff/OpenStackDashboardSrc/horizon/openstack-dashboard/.dashboard-venv/src/quantum (from dependency_links.txt)" (including the typo in "directores") | 16:36 |
| *** hadrian has joined #openstack | 16:38 | |
| *** Badri has quit IRC | 16:40 | |
| *** mikeyp has joined #openstack | 16:41 | |
| vidd | Kiall, you here? | 16:41 |
| vidd | BuZZ-T, the first "error" is your using the venv | 16:42 |
| vidd | the second is, since your using the venv, quantum is not getting into your venv | 16:43 |
| *** jakedahn has joined #openstack | 16:43 | |
| vidd | BuZZ-T, is there a reason you dont use Kiall 's repos? | 16:43 |
| vidd | BuZZ-T, https://github.com/managedit/openstack-setup | 16:44 |
| *** Turicas has joined #openstack | 16:44 | |
| *** nyeates has joined #openstack | 16:45 | |
| *** Razique has quit IRC | 16:45 | |
| Glacee | Swift question: Lets say if your cluster reach the point of maximum drives ( 100 partitions per drive).. your pretty much screwed right? any good experience of rings resizing in prod? | 16:47 |
| *** nphase has quit IRC | 16:49 | |
| *** vdo has joined #openstack | 16:50 | |
| *** negronjl has joined #openstack | 16:50 | |
| koolhead17 | vidd, am still stuck at same place :( | 16:51 |
| vidd | koolhead17, dont feel bad...so am i | 16:51 |
| vidd | infact...im doing a clean install right now | 16:52 |
| *** jaypipes has quit IRC | 16:52 | |
| Kiall | vidd, yea.. i'm around | 16:54 |
| BuZZ-T | vidd: i just followed the instructions from http://docs.openstack.org/diablo/openstack-compute/admin/content/install-dashboard.html | 16:54 |
| vidd | koolhead17, where are you stuck at exactly | 16:54 |
| *** jakedahn has quit IRC | 16:54 | |
| *** jakedahn has joined #openstack | 16:54 | |
| koolhead17 | vidd, let me pastebin exact comamnds i used | 16:54 |
| koolhead17 | hey Kiall | 16:54 |
| vidd | BuZZ-T, you followed the instructions for diablo but you installed essex dashboard | 16:54 |
| Kiall | n88, ping | 16:55 |
| *** katkee has quit IRC | 16:55 | |
| BuZZ-T | i see... | 16:55 |
| vidd | Kiall, i was going to ask you for some assistance getting glance and keystone to work together...but i have already formatted and started a re-install | 16:56 |
| Kiall | n88, you mentioned that my dashboard packages depend on nova-compute a while back .. As near as I can tell, they dont! | 16:56 |
| Kiall | vidd, lol | 16:56 |
| BuZZ-T | so what's the better way? Installing the dashboard for diablo or using an instruction for essex? | 16:56 |
| Kiall | may I suggest trying all this in a VM that you can snapshot? ;) | 16:56 |
| Kiall | You must be on install #250 at this stage | 16:56 |
| vidd | Kiall, im going to follow your stripts line by line and see if the result is different | 16:57 |
| vidd | Kiall, ive already had to replace the hard drive once...i think i wore a hole in the platter of the old one =] | 16:57 |
| Kiall | My keystone+glance+dashboard+nova "scripts" (up to the network command) should work for any network type ;) | 16:58 |
| *** nphase has joined #openstack | 16:58 | |
| *** maplebed has joined #openstack | 16:58 | |
| *** obino has joined #openstack | 16:58 | |
| vidd | Kiall, you do need to add a line to your README about installing that python tool to get a PPA added though | 16:59 |
| koolhead17 | vidd, http://paste.ubuntu.com/741366/ this is what i followed | 17:00 |
| Kiall | vidd, https://github.com/managedit/openstack-setup/commit/ee32a45266cc83bc8e6fc2582114cce63c72401c | 17:01 |
| vidd | koolhead17, tools/with_venv.sh dashboard/manage.py syncdb | 17:01 |
| *** pradeep has quit IRC | 17:01 | |
| vidd | there you go using the venv | 17:01 |
| koolhead17 | vidd, yes am using virtualenv | 17:01 |
| vidd | then i cant help you | 17:02 |
| *** TheOsprey has quit IRC | 17:03 | |
| vidd | the venv isolated openstack dashboard from django-openstack and i dont know how to get around that | 17:03 |
| *** kaigan has left #openstack | 17:04 | |
| vidd | if you lose the venv you should lose the issue | 17:04 |
| *** RobertLaptop has joined #openstack | 17:05 | |
| *** rnorwood has joined #openstack | 17:05 | |
| koolhead17 | vidd, does it mean this isolation or code change has happened recently? i was able to setup the dashboard same way earlier | 17:05 |
| vidd | no...im saying there is a way around this but i dont know what it is | 17:06 |
| *** StarX has joined #openstack | 17:06 | |
| *** RobertLaptop has left #openstack | 17:07 | |
| vidd | your missing some step that i never took | 17:07 |
| *** aliguori has quit IRC | 17:07 | |
| vidd | koolhead17, is there some reason you MUST use isolation? | 17:08 |
| *** maplebed has left #openstack | 17:10 | |
| *** hsmod has joined #openstack | 17:11 | |
| Kiall | vidd, https://github.com/managedit/openstack-setup/commit/9a1259346691bb677bd25e9f9560cae9ff89a9d7 | 17:12 |
| *** lvaughn_ has quit IRC | 17:13 | |
| *** lvaughn has joined #openstack | 17:13 | |
| *** dolphm has joined #openstack | 17:18 | |
| *** phschwartz has quit IRC | 17:19 | |
| *** helfrez has joined #openstack | 17:19 | |
| *** lvaughn_ has joined #openstack | 17:22 | |
| *** mencken has joined #openstack | 17:23 | |
| *** mrevell has joined #openstack | 17:25 | |
| *** mrevell has joined #openstack | 17:26 | |
| *** lvaughn has quit IRC | 17:26 | |
| vidd | Kiall, you may want to review my "basics" script for inclusion | 17:30 |
| Kiall | lets see? | 17:31 |
| *** phschwartz has joined #openstack | 17:31 | |
| vidd | Kiall, it does things like tell ya "hey! you dont have apache...." | 17:31 |
| vidd | https://github.com/vidd/vidd/blob/master/basics.sh | 17:32 |
| Kiall | But apache is listed as a dependency of openstack-dashboard ;) | 17:32 |
| Kiall | Also - https://github.com/managedit/openstack-setup/commit/caa599994cbd8c72a88296fa3052757c4d6fb2d8 | 17:32 |
| vidd | the very first thing your script did was tell me it could talk to mysql =\ | 17:33 |
| Kiall | could? | 17:33 |
| Kiall | or couldnt? | 17:33 |
| vidd | could not | 17:33 |
| Kiall | https://github.com/managedit/openstack-setup/blob/master/keystone.sh | 17:33 |
| Kiall | The first thing is does is install what it needs... | 17:34 |
| *** juddm has joined #openstack | 17:34 | |
| *** aliguori has joined #openstack | 17:35 | |
| vidd | right...but it does not set mysql to listen to 0.0.0.0 | 17:35 |
| Kiall | Oh .. | 17:35 |
| hsmod | is there a channel dedicated to swift? | 17:35 |
| Spirilis | don't think there is.. | 17:36 |
| *** cmasseraf has joined #openstack | 17:36 | |
| *** vdo has quit IRC | 17:38 | |
| *** Satya has quit IRC | 17:39 | |
| Kiall | vidd, does that basics.sh not screw with the ntp.conf? ;) | 17:40 |
| Kiall | it looks like you have used n\ rather than \n | 17:40 |
| hsmod | following http://swift.openstack.org/howto_installmultinode.html, when i do the first test curl cmd curl -k -v -H 'X-Storage-User: system:root' -H 'X-Storage-Pass: testpass' https://$PROXY_LOCAL_NET_IP:8080/auth/v1.0, i get: | 17:40 |
| hsmod | Exception: Could not create account AUTH_system for user system:root | 17:40 |
| vidd | no...its right | 17:41 |
| vidd | let me triplecheck | 17:41 |
| hsmod | i checked my /srv/node dir is correctly owned by swift:swift (and its subdirs) | 17:41 |
| vidd | it does it correctly | 17:41 |
| hsmod | /var/log/syslog gives me no indication of any problems. anyone have any idea what it might be? | 17:41 |
| Kiall | vidd, https://github.com/managedit/openstack-setup/commit/1f0f96a69289d04a78cf3cfb8bc1c62b01911ee0 | 17:42 |
| Spirilis | hsmod: you got the account-server running on all the nodes? | 17:42 |
| hsmod | Spirilis, i'm using tempauth | 17:42 |
| Spirilis | ah gotcha | 17:42 |
| hsmod | as instructed in that url | 17:42 |
| hsmod | should i just go to swauth instead? | 17:43 |
| vidd | Kiall, n\ is bash reserved work for "new line" | 17:43 |
| Kiall | ah.. news to me ;) | 17:43 |
| *** vernhart has joined #openstack | 17:43 | |
| Spirilis | though it sounds like account-server is still needed | 17:43 |
| Spirilis | even with tempauth | 17:44 |
| hsmod | ok. i'll try that | 17:44 |
| hsmod | thx | 17:44 |
| Spirilis | or at least that document tells you to write its config file on the storage nodes and start the account-* services... | 17:44 |
| *** TheOsprey has joined #openstack | 17:45 | |
| hsmod | yes but the account-container services are not really the auth | 17:45 |
| Spirilis | yeah | 17:45 |
| hsmod | accont-container is the ring manager for the account-builder rings if i understand correclty | 17:45 |
| *** jog0 has joined #openstack | 17:46 | |
| Spirilis | the account ring does have to contain the 'system' account you're creating though right? | 17:46 |
| hsmod | right, but isn't that transferred when i scp the .gz files over? | 17:46 |
| hsmod | i'll try anything though... i'm just trying to clarify which file i should try copying to storage nodes. | 17:47 |
| hsmod | :) | 17:47 |
| Spirilis | naw the .gz files are rings, meaning, definitions for where data objects created in that "ring" should go | 17:47 |
| Spirilis | that whole ring concept took me like 2 weeks to finally understand when I first installed swift | 17:47 |
| hsmod | oh ok, so which files should i ensure are transferred to storage nodes? | 17:47 |
| hsmod | yes tell me about it :) rings are not straightforward | 17:48 |
| Spirilis | just make sure all the .gz files are transferred and all the account, container & object services are running | 17:48 |
| hsmod | indeed, i ensured all those things | 17:48 |
| Spirilis | could be something else going wrong and that'd be outside my knowledge :) | 17:48 |
| hsmod | you mentioned something about a config | 17:48 |
| Spirilis | yeah just talking about the /etc/swift/account-server.conf file that you created if you followed that webpage's steps | 17:48 |
| Spirilis | what they provided should be correct I believe | 17:48 |
| hsmod | yes i did that | 17:49 |
| hsmod | it doesn't have any info pertinent to the tempauth acccount tho, so i figured wasn't the issue | 17:49 |
| Spirilis | yeah probably not | 17:49 |
| hsmod | the tempauth creds appear to be in the proxy.conf | 17:49 |
| hsmod | which i moved over to storage nodes just for kicks | 17:49 |
| hsmod | i'll just bite the bullet and try swauth or keystone. which do u recommend for proof of concept? | 17:50 |
| hsmod | i heard keystone is not fun | 17:50 |
| *** jaypipes has joined #openstack | 17:50 | |
| Spirilis | swauth wasn't too hard for me IMO | 17:50 |
| hsmod | ok cool, i'll try that out then | 17:50 |
| Spirilis | and our POC runs it for now | 17:50 |
| hsmod | thanks for your help! | 17:51 |
| Spirilis | np | 17:51 |
| *** dprince has quit IRC | 17:51 | |
| vidd | Kiall, look at this:http://dpaste.com/658184/ | 17:51 |
| Kiall | vidd, yea obtain-token.sh failed | 17:51 |
| Kiall | try the curl command by hand | 17:51 |
| vidd | and ppl wonder why i keep my hair cut less then 1/4 inch =] | 17:51 |
| vidd | this is the exact same issue i had with my setup scripts with glance | 17:52 |
| *** jog0 has quit IRC | 17:53 | |
| vidd | i dont understand how to do "curl command by hand" | 17:53 |
| *** vernhart has quit IRC | 17:53 | |
| Kiall | https://github.com/managedit/openstack-setup/blob/master/obtain-token.sh | 17:53 |
| *** Ryan_Lane has joined #openstack | 17:53 | |
| Kiall | bwteen L8 and L9 add 2x lines: | 17:54 |
| Kiall | echo $RAW_TOKEN | 17:54 |
| Kiall | exit | 17:54 |
| *** dysinger has joined #openstack | 17:55 | |
| vidd | raw token = NULL | 17:55 |
| Kiall | then keystone is failing for some reason | 17:56 |
| Glacee | Swift question: Lets say if your cluster reach the point of maximum drives ( 100 partitions per drive).. your pretty much screwed right? any good experience of rings resizing in prod? | 17:58 |
| *** Ryan_Lane has quit IRC | 17:58 | |
| *** mdomsch_ has joined #openstack | 17:58 | |
| *** mdomsch has quit IRC | 17:59 | |
| vidd | i did http://paste.ubuntu.com/741430/ | 18:00 |
| vidd | and i get the first 2 but no raw token | 18:00 |
| vidd | Kiall, | 18:00 |
| Kiall | RAW_TOKEN should be whatever HTTP respose keystone is giving | 18:01 |
| Kiall | back | 18:01 |
| Kiall | so, If its not giving something, keystone is not working for some reason? | 18:02 |
| vidd | is it possible im missing this json thing | 18:02 |
| koolhead17 | is quantum a must requirement/dependency for dashboard | 18:03 |
| kbringard | don't think so | 18:03 |
| Kiall | koolhead17, no | 18:03 |
| vidd | koolhead17, no.... | 18:03 |
| kbringard | keystone is, though | 18:03 |
| kbringard | at least for any recent rev | 18:04 |
| vidd | set it to "false" in your local_settings.py | 18:04 |
| *** jdurgin has joined #openstack | 18:04 | |
| *** vernhart has joined #openstack | 18:05 | |
| *** Shentonfreude has quit IRC | 18:05 | |
| vidd | Kiall, so...your keystone is broken...not my script [i feel MUCH better] | 18:06 |
| *** mrevell has quit IRC | 18:06 | |
| Kiall | vidd, go on? | 18:06 |
| vidd | your keystone is not working | 18:06 |
| koolhead17 | vidd, now i think http://docs.openstack.org/diablo/openstack-compute/admin/content/install-dashboard.html this needs to be modified | 18:07 |
| *** dprince has joined #openstack | 18:07 | |
| vidd | i didnt use my script at all this time so its not my script that did it | 18:07 |
| Kiall | vidd, okay .. but how is it not working? | 18:07 |
| koolhead17 | i spent lots many hours in setting up dashboard and no luck :) | 18:07 |
| vidd | Kiall, i dont understand how it works when it does work....how am i supposed to know how it isnt working | 18:08 |
| Kiall | lol | 18:08 |
| Kiall | Okay, I'll tell you what.. I'll test it on a clean VM ;) | 18:08 |
| vidd | anyway...i say its broken because it does not retern a token | 18:08 |
| vidd | which means glance is not going to work | 18:08 |
| *** mjfork has joined #openstack | 18:09 | |
| Kiall | well, if keystone-test.sh worked earlier.. | 18:09 |
| Kiall | then, it worked earlier ;) | 18:09 |
| vidd | keystone-test.sh FAILED before...even though it said it succeded | 18:09 |
| Kiall | lol, how did it fail? Or are you assuming it failed? | 18:10 |
| *** nerens has quit IRC | 18:10 | |
| Kiall | n88, ping | 18:10 |
| Kiall | you just got my keystone packages working right? | 18:10 |
| vidd | 8 lines of error code followed by a single line "Success" does not assume me things are well in the world =\ | 18:10 |
| Kiall | Then why did you move on -_- | 18:11 |
| vidd | i didnt | 18:11 |
| *** CaptTofu has joined #openstack | 18:11 | |
| vidd | im at " ./keystone-test.sh | 18:11 |
| vidd | " | 18:11 |
| vidd | epic failure | 18:11 |
| Kiall | vidd, I'm installing a VM on esx at the mo, give me 30 mins... | 18:12 |
| vidd | take your time | 18:12 |
| *** medberry is now known as mka | 18:13 | |
| vidd | ill be over here in the corner crying in my coffee and banging my head against the wall =] | 18:14 |
| *** mka is now known as medberry | 18:14 | |
| *** adjohn has joined #openstack | 18:14 | |
| *** cmasseraf has quit IRC | 18:15 | |
| *** katkee has joined #openstack | 18:16 | |
| *** mdomsch_ has quit IRC | 18:16 | |
| *** GheRivero_ has joined #openstack | 18:17 | |
| *** cmasseraf has joined #openstack | 18:17 | |
| *** ohnoimdead has joined #openstack | 18:18 | |
| *** Shentonfreude has joined #openstack | 18:18 | |
| hsmod | Spirilis, so i installed swauth, made the necessary configs to proxy conf file, fired up the proxy and so when I try to use any of the swauth cmds to initialize the backend or even create a user, it just hangs indefinitely. telnet to the port (8080) to do a GET / also hangs indefinitely. | 18:20 |
| dolphm | vidd: Kiall: what's the keystone issue ya'll are having? | 18:20 |
| hsmod | Spirilis, you ever seen that before? | 18:20 |
| *** rustam has quit IRC | 18:20 | |
| *** woleium has joined #openstack | 18:21 | |
| vidd | dolphm, we are using Kiall 's ppa and keystone wont return a token | 18:21 |
| Kiall | vidd, it does fore n88 and I ;) | 18:21 |
| Kiall | for* | 18:21 |
| Spirilis | hsmod: that sounds AWFULLY familiar... trying to remember what it was... | 18:21 |
| dolphm | vidd: stable/diablo? | 18:21 |
| Kiall | dolphm, yea.. my PPA packages are stable/diablo | 18:22 |
| Kiall | they work for me, and a bunch of others.. he has something setup wrong! I'm *sure* of it | 18:22 |
| dolphm | Kiall: do you have a relevant paste or something i can look at? | 18:22 |
| dolphm | vidd: ^^ | 18:22 |
| Kiall | Yea.. https://github.com/managedit/openstack-setup/ | 18:23 |
| vidd | Kiall, i cant possibly have something set up wrong =\ its a fresh install i ahve not set up anything ! | 18:23 |
| *** negronjl has quit IRC | 18:23 | |
| Kiall | dolphm, there is a keystone-paste template in there.. | 18:23 |
| Kiall | and data template | 18:23 |
| vidd | is this json thing something that needs to be installed and im somehow not getting it? | 18:24 |
| dolphm | Kiall: let me fire up a server then | 18:24 |
| Spirilis | hsmod: can't recall the exact reason I had that, but can you make sure each box's hostname resolves to its correct IP (with hostname -i)? think it might've been related to that... | 18:24 |
| vidd | im doing a stock ubuntu 11.10 installer from usb stick | 18:25 |
| *** negronjl has joined #openstack | 18:25 | |
| hsmod | Spirilis, ok | 18:25 |
| hsmod | thx again for the tip , i'll keep looking | 18:25 |
| vidd | maybe ill burn down another install and test that devstack thing =] | 18:25 |
| Kiall | lol | 18:25 |
| *** jog0 has joined #openstack | 18:25 | |
| *** dirkx_ has quit IRC | 18:26 | |
| Kiall | vidd, stop. All that you get from re-installing is a waste of time.. | 18:26 |
| Kiall | You're missing something. And, Since you dont know what it is, You're going to be missing it again next time. | 18:26 |
| vidd | thats why im asking...is this json thing the curl is crying about something i can install | 18:27 |
| *** magg has joined #openstack | 18:28 | |
| Spirilis | hsmod: also anything on /var/log/syslog on the proxy host? | 18:28 |
| hsmod | nothing :( | 18:28 |
| magg | if im using --multi_host=T do i need to put this in the nova.conf | 18:28 |
| Spirilis | unfortunately I didn't jot down why swauth kept doing that but I remember it being a roadblock for a little while | 18:28 |
| hsmod | just that it started up the children | 18:29 |
| hsmod | lots of roadblocks, that's for sure :) | 18:29 |
| Kiall | vidd, its not that, since when you exit'ed before the JSON parse, it had nothing to parse | 18:30 |
| Spirilis | ok one thing I recall adding to my proxy-server.conf is under the [filter:swauth] section, "default_swift_cluster = local#https://<proxy-node-hostname>/v1" ... if you're just using http with port 8080 then make that http://<proxy-node-hostname>:8080/v1 | 18:31 |
| Spirilis | and additionally under [app:auth-server], a "default_cluster_url = https://<proxy-node>/v1" (make similar changes for your http-port-8080 config) | 18:32 |
| dolphm | Kiall: which OS are you intending this for? | 18:32 |
| dolphm | Kiall: 11.04? 11.10? | 18:32 |
| Kiall | 11.10 | 18:32 |
| Spirilis | otherwise those sections also use the same "super_admin_key" with the same password of course | 18:33 |
| Spirilis | and it looks like that "super_admin_key" directive is only found in the [filter:swauth] and [app:auth-server] sections | 18:34 |
| dolphm | Kiall: i can't test it right now then :/ | 18:34 |
| *** perestre1ka has quit IRC | 18:34 | |
| dolphm | Kiall: well, maybe i can.. | 18:35 |
| Kiall | dolphm, no hassle .. I'm installing a VM at the mo to test this with | 18:35 |
| magg | if im using --multi_host=T do i need to put this in the nova.conf??? help | 18:35 |
| *** PiotrSikora_ is now known as PiotrSikora | 18:38 | |
| hsmod | Spirilis, i don't have an '[app:auth-server]' section at all in my proxy-server.conf | 18:38 |
| hsmod | maybe that is my problem. i got my proxy-server.conf from http://swift.openstack.org/howto_installmultinode.html | 18:39 |
| Spirilis | hsmod: [app:auth-server], underneath that "use = egg:swift#auth", under that "default_cluster_url = https://<proxynodehostname>/v1", under that "super_admin_key = <my super admin key password>" | 18:39 |
| hsmod | although i don't see that section here either http://docs.openstack.org/diablo/openstack-object-storage/admin/content/installing-and-configuring-the-proxy-node.html | 18:39 |
| hsmod | ok thx i'll try that | 18:39 |
| hsmod | i wonder why they omitted that from docs | 18:40 |
| Spirilis | ok that is weird | 18:40 |
| Spirilis | yeah | 18:40 |
| hsmod | oh that goes into auth-server.conf | 18:40 |
| *** localhost has quit IRC | 18:41 | |
| Spirilis | erm hm, I don't have an auth-server.conf and that's at the end of my proxy-server.conf | 18:41 |
| *** perestrelka has joined #openstack | 18:41 | |
| hsmod | strange! | 18:41 |
| Kiall | vidd, VM installed.. Will snapshot it and make sure those instructions work as advertised ;) | 18:42 |
| hsmod | ya it only exists here http://swift.openstack.org/1.2/howto_installmultinode.html (old version of doc)? doesn't exist here http://swift.openstack.org/howto_installmultinode.html | 18:42 |
| hsmod | anyway, adding now | 18:42 |
| vidd | hehe | 18:42 |
| Spirilis | yeah, you're retracing all my old steps lol | 18:42 |
| *** localhost has joined #openstack | 18:42 | |
| *** carlp has quit IRC | 18:42 | |
| hsmod | did you have to do a swift-init auth start? | 18:42 |
| Spirilis | I was doing google searches all over the place looking for example configs until I stumbled across something that worked | 18:42 |
| hsmod | or do u have to startup swauth at all? i didn't do any of that | 18:43 |
| *** dubsquared has joined #openstack | 18:43 | |
| *** gyee has joined #openstack | 18:43 | |
| Spirilis | no | 18:43 |
| Spirilis | the proxy node only needed swift-init proxy-server start | 18:43 |
| Spirilis | and storage nodes needed the account-*, container-* and object-* servers various start cmds | 18:43 |
| Spirilis | nothing about auth-server | 18:43 |
| Spirilis | I think it's built into the proxy server, that's why it's in proxy-server.conf | 18:44 |
| *** carlp has joined #openstack | 18:45 | |
| magg | i need help with a multi_host configuration | 18:45 |
| *** scottjg has joined #openstack | 18:45 | |
| hsmod | no luck. | 18:46 |
| *** whaley has quit IRC | 18:48 | |
| Spirilis | k then I don't know what's wrong :) | 18:48 |
| hsmod | thanks for trying :) | 18:49 |
| Spirilis | np | 18:49 |
| vidd | Kiall, the curl is now working | 18:53 |
| *** hsmod has quit IRC | 18:53 | |
| *** dendro-afk is now known as dendrobates | 18:53 | |
| Kiall | what did you change? | 18:53 |
| vidd | idk | 18:54 |
| koolhead17 | ok i had some luck. now the dashborad says "Could not import dashboard.views. Error was: No module named Crypto.Cipher | 18:54 |
| koolhead17 | " | 18:54 |
| *** StarX has quit IRC | 18:55 | |
| Kiall | koolhead17, apt-get install python-crypto maybe | 18:55 |
| vidd | i did a clean install with no LAMP | 18:55 |
| vidd | and added none of my normal packages | 18:55 |
| vidd | no mysql, no apache | 18:56 |
| vidd | so something on my basics list or in the lamp install is messing it up? | 18:56 |
| vidd | or was there a typo somewhre | 18:56 |
| *** katkee has quit IRC | 18:56 | |
| Kiall | vidd, give me 20 mins.. I'll be back with 100% working scripts ;) | 18:57 |
| *** ahale has joined #openstack | 18:57 | |
| vidd | Kiall, you might find it interesting that there are currently NO endpoints in the keystone database | 18:58 |
| *** rnorwood has quit IRC | 18:58 | |
| *** hsmod has joined #openstack | 18:58 | |
| *** darraghb has quit IRC | 19:00 | |
| *** swill has left #openstack | 19:02 | |
| vidd | Kiall, glance.sh is missing httplib2 | 19:02 |
| *** nacx has quit IRC | 19:03 | |
| * koolhead17 wonders why pycrypto was not added in pip-requires file :( | 19:03 | |
| *** rnorwood has joined #openstack | 19:03 | |
| vidd | Kiall, python-httplib2 | 19:04 |
| Kiall | Humm - really? | 19:04 |
| Kiall | I wonder which specific package needs that.. | 19:04 |
| vidd | /usr/lib/python2.7/dist-packages/glance/common/auth.py", line 33, in <module> | 19:05 |
| vidd | import httplib2 | 19:05 |
| vidd | i suppose that is the api? | 19:05 |
| vidd | glance-manage =] | 19:06 |
| Kiall | I'll figure it out later, for now just install it ;) | 19:06 |
| vidd | already done...just letting you knwp | 19:06 |
| Kiall | k | 19:06 |
| *** dev_sa has joined #openstack | 19:06 | |
| *** dev_sa has left #openstack | 19:07 | |
| *** whaley has joined #openstack | 19:07 | |
| *** dirkx_ has joined #openstack | 19:07 | |
| *** ivano has joined #openstack | 19:08 | |
| ivano | ciao | 19:08 |
| *** imsplitbit has quit IRC | 19:08 | |
| *** ivano has quit IRC | 19:09 | |
| *** ivano has joined #openstack | 19:09 | |
| Kiall | vidd, dooh | 19:10 |
| Kiall | somehow I've just done this VM with the ubuntu packages -_- | 19:10 |
| Kiall | thank god for snapshots ;) | 19:11 |
| hsmod | Spirilis, i was finally able to get some response by doing : swauth-prep -K swauthkey -A https://192.168.0.50/auth | 19:11 |
| vidd | uhhh whoops? | 19:11 |
| hsmod | go figure | 19:11 |
| hsmod | correction, -A https://192.168.0.50:8080/auth | 19:11 |
| vidd | Kiall, apt-get auto-remove --purge? | 19:11 |
| Kiall | vidd, already back at a clean VM.. | 19:12 |
| *** kerouac has quit IRC | 19:12 | |
| *** katkee has joined #openstack | 19:12 | |
| *** binbash_ has quit IRC | 19:13 | |
| *** ryan_fox1985 has quit IRC | 19:13 | |
| *** webx has quit IRC | 19:13 | |
| *** taihen has quit IRC | 19:13 | |
| *** vishy has quit IRC | 19:13 | |
| *** gohko_nao has quit IRC | 19:13 | |
| *** dmarkey has quit IRC | 19:13 | |
| *** laurensell has quit IRC | 19:13 | |
| *** guitarzan has quit IRC | 19:13 | |
| *** AntoniHP has quit IRC | 19:13 | |
| *** aimka has quit IRC | 19:13 | |
| *** comstud has quit IRC | 19:13 | |
| *** Archeron has quit IRC | 19:13 | |
| *** root_ has quit IRC | 19:13 | |
| *** dannf has quit IRC | 19:13 | |
| *** Aurelgadjo has quit IRC | 19:13 | |
| *** cloud0_ has quit IRC | 19:13 | |
| *** cloud0_ has joined #openstack | 19:13 | |
| *** guitarzan has joined #openstack | 19:13 | |
| *** root_ has joined #openstack | 19:13 | |
| *** AntoniHP has joined #openstack | 19:13 | |
| *** taihen has joined #openstack | 19:13 | |
| *** binbash_ has joined #openstack | 19:13 | |
| *** vishy has joined #openstack | 19:13 | |
| *** dannf has joined #openstack | 19:13 | |
| *** dmarkey has joined #openstack | 19:13 | |
| *** comstud has joined #openstack | 19:13 | |
| *** webx has joined #openstack | 19:13 | |
| *** ryan_fox1985 has joined #openstack | 19:13 | |
| *** Archeron has joined #openstack | 19:13 | |
| *** gohko_nao has joined #openstack | 19:13 | |
| *** laurensell has joined #openstack | 19:14 | |
| *** aimka has joined #openstack | 19:14 | |
| *** magg has left #openstack | 19:14 | |
| vidd | Kiall, this is interesting..... ./keystone-test.sh | 19:15 |
| vidd | returns a token but ./glance-upload-oneiric.sh | 19:15 |
| vidd | fails as "unauthorized | 19:15 |
| *** haji has joined #openstack | 19:15 | |
| haji | hello | 19:15 |
| koolhead17 | booooh. dashboard does need quantum pkg | 19:15 |
| Kiall | vidd, few mins ;) | 19:16 |
| Kiall | koolhead17, no it doesnt -_- | 19:16 |
| koolhead17 | Kiall, Could not import dashboard.views. Error was: No module named quantum | 19:16 |
| haji | so nova-network gets a XXX on my controller node | 19:16 |
| koolhead17 | i removed that quantum git path from pip-requires | 19:16 |
| *** timo_ has joined #openstack | 19:17 | |
| Kiall | koolhead17, odd.. mayber your using the essex dashboard? | 19:17 |
| *** timo_ is now known as Guest78382 | 19:17 | |
| vidd | koolhead17, did you set the quatum to "false" in the local.py? | 19:17 |
| koolhead17 | vidd, yes i did that | 19:17 |
| Guest78382 | good afternoon all! Im having some issues with my swift test cluster, getting object6000 no found from rsync. Can anyone help me out? | 19:18 |
| koolhead17 | Kiall, git clone https://github.com/4P/horizon | 19:18 |
| koolhead17 | is that wrong ? | 19:18 |
| Kiall | Yes, For one.. Its https://github.com/openstack/horizon ;) | 19:18 |
| vidd | koolhead17, that will get you the "master" vesion | 19:18 |
| Kiall | And both of those commands on their own will give you the essex version | 19:19 |
| haji | and nova-network.log doesnt say anything | 19:19 |
| haji | help | 19:19 |
| koolhead17 | so what should is use instead Kiall ? | 19:19 |
| *** dysinger has quit IRC | 19:20 | |
| Kiall | your using diablo right? You should use the diablo branch of those repos | 19:20 |
| *** jog0 has quit IRC | 19:20 | |
| *** jog0_ has joined #openstack | 19:20 | |
| vidd | koolhead17, git clone https://github.com/openstack/horizon.git horizon && horizon && git checkout stable/diablo | 19:20 |
| *** dysinger has joined #openstack | 19:20 | |
| koolhead17 | Kiall, yes diablo | 19:20 |
| vidd | or use kiall's repos =] | 19:20 |
| haji | :( | 19:20 |
| Kiall | lol | 19:20 |
| Kiall | haji, no errors? | 19:21 |
| haji | nop | 19:21 |
| Kiall | and NTP is okay between all the servers? | 19:21 |
| koolhead17 | Kiall, vidd thanks let me try that | 19:21 |
| haji | kiall: i installed nova-network on both nodes | 19:21 |
| haji | like u said | 19:21 |
| Kiall | haji, did those VLAN commands work BTW? (I replied to your mailing list message) | 19:21 |
| *** jog0_ has quit IRC | 19:21 | |
| haji | yeah i say it thanks | 19:21 |
| koolhead17 | http://docs.openstack.org/diablo/openstack-compute/admin/content/installing-openstack-dashboard.html am filing a bug and assigning to myself. | 19:21 |
| haji | they worked | 19:21 |
| *** jog0 has joined #openstack | 19:21 | |
| praefect | I have issues with lock files race condition and it's driving me nuts!!!! anyone ever seen something like that??? | 19:21 |
| Kiall | didnt notice a respose, openstack flooded my inbox today ;) | 19:22 |
| haji | i saw it*** | 19:22 |
| haji | u were the first to reply | 19:22 |
| *** stewart has joined #openstack | 19:22 | |
| vidd | thats cuzz Kiall rocks =] | 19:22 |
| Kiall | lol | 19:22 |
| vidd | his statue is in the mail somewhere.... | 19:22 |
| *** dendrobates is now known as dendro-afk | 19:22 | |
| haji | hehehe | 19:24 |
| Kiall | vidd, weird.. having thr same issue with glance.. | 19:24 |
| Kiall | but just restarting them fixed it | 19:24 |
| Kiall | even thought I did already restart them -_- | 19:24 |
| vidd | restarting....glance? keystone? | 19:24 |
| haji | but, yeah nova-network is XXX on the controller... and i dont have any br or vlan interfaces on the other node | 19:24 |
| vidd | the server? | 19:24 |
| Kiall | glance | 19:25 |
| Kiall | api and reg | 19:25 |
| haji | is this because im using multi node T? | 19:25 |
| vidd | haji, did you state host ip on each machine as the controller ip? | 19:25 |
| *** cmasseraf has quit IRC | 19:25 | |
| *** scottjg has quit IRC | 19:26 | |
| haji | mmm | 19:26 |
| haji | yeah... | 19:26 |
| haji | well in the nova.conf almost all IP have the controllers ip | 19:27 |
| zykes- | wazzup vidd ? | 19:27 |
| jeh | anyone using nginx to load balance nova-api? | 19:27 |
| koolhead17 | vidd, horizon: command not found | 19:27 |
| vidd | haji, there should also be a --my_ip flag set to the server ip the file resides on | 19:27 |
| koolhead17 | when i execute git clone https://github.com/openstack/horizon.git horizon && horizon && git checkout stable/diablo | 19:28 |
| koolhead17 | zykes-, where have you been :D | 19:28 |
| *** negronjl has quit IRC | 19:28 | |
| haji | --my_ip=10.10.10.2 on the controller? and --my_ip=10.10.10.3 on the compute? | 19:29 |
| vidd | sorry koolhead17 missed the "cd" git clone https://github.com/openstack/horizon.git horizon && cd horizon && git checkout stable/diablo | 19:30 |
| haji | well ok | 19:30 |
| haji | done | 19:30 |
| vidd | hello zykes- | 19:30 |
| vidd | just banging my head against a wall =] | 19:31 |
| koolhead17 | vidd, so am i | 19:31 |
| koolhead17 | :P | 19:31 |
| *** timolow has joined #openstack | 19:32 | |
| vidd | the ppl with the white coats and the pleasent smiles are coming to visit me later with the intersting night-robe and the sleeping room with the cusioned walls....=] | 19:32 |
| haji | maybe the two nova-networks are conflicting | 19:32 |
| *** negronjl has joined #openstack | 19:32 | |
| haji | with each other? | 19:32 |
| vidd | [in other words they're going to be dragging me away in a stright-jacket to a padded room] | 19:33 |
| koolhead17 | Kiall, even in diablo/stable i see -e git+https://github.com/openstack/quantum.git#egg=quantum | 19:33 |
| koolhead17 | inside pip-require . let me do it all again :p | 19:33 |
| timolow | howdy, im getting errors with my swift implementation. anyone come across this error with rsync before? 2011/11/17 19:33:20 [14561] connect from i-00006471.novalocal (10.4.1.5) | 19:34 |
| timolow | 2011/11/17 19:33:20 [14561] unknown module 'object6000' tried from i-00006471.novalocal (10.4.1.5) | 19:34 |
| Kiall | koolhead17, yea, dashboard can make use of quantum | 19:34 |
| Kiall | but does not require it | 19:34 |
| haji | kiall i need your nova.confs plz | 19:34 |
| *** dendro-afk is now known as dendrobates | 19:34 | |
| Kiall | haji, https://github.com/managedit/openstack-setup/blob/master/nova.conf.tmpl | 19:34 |
| *** fabiand__ has joined #openstack | 19:35 | |
| vidd | Kiall, still no dice on glance | 19:35 |
| Kiall | vidd, give me a few ;) | 19:36 |
| haji | kiall so i need to put the same file on the compute nodes | 19:36 |
| Kiall | haji, yea.. they should all be the same pretty much | 19:36 |
| vidd | Kiall, im sorry...i thought you said it worked for you and i just had to restart | 19:36 |
| haji | i dont need extra commands for the multi node T config | 19:37 |
| Kiall | oh .. sorry, still checking everything ;) | 19:37 |
| vidd | i know you can only do 15 things at once =] | 19:38 |
| Kiall | Well, Waiting for the oneiric and lucid image downloads ;) | 19:38 |
| *** Ryan_Lane has joined #openstack | 19:39 | |
| haji | and on the compute nodes i install nova-compute nova-network and mysql-client only? | 19:39 |
| haji | do i need nova-api? | 19:39 |
| *** blamar has quit IRC | 19:39 | |
| Kiall | I put nova-api on, but its not needed at all | 19:39 |
| *** blamar has joined #openstack | 19:39 | |
| *** nerens has joined #openstack | 19:39 | |
| *** reiddraper has joined #openstack | 19:40 | |
| *** mirrorbox has quit IRC | 19:43 | |
| *** mirrorbox has joined #openstack | 19:43 | |
| *** cmasseraf has joined #openstack | 19:43 | |
| haji | ook | 19:45 |
| *** anotherjesse has joined #openstack | 19:46 | |
| Kiall | vidd, it looks like `keystone-manage token add` is failing | 19:47 |
| *** Guest78382 has quit IRC | 19:47 | |
| vidd | vunderbar | 19:48 |
| vidd | so it is NOT me =] | 19:48 |
| vidd | i feel better now =] | 19:49 |
| vidd | Kiall, do you have endpoints in your keystone database? | 19:50 |
| *** PeteDaGuru has quit IRC | 19:50 | |
| *** PeteDaGuru1 has joined #openstack | 19:50 | |
| *** imsplitbit has joined #openstack | 19:50 | |
| Kiall | no | 19:51 |
| *** negronjl has quit IRC | 19:51 | |
| Kiall | Do have endpoint_templates | 19:51 |
| vidd | could that be the issue? | 19:51 |
| Kiall | Nope.. Dont think so | 19:51 |
| Kiall | My live stuff has an empty endpoints table | 19:51 |
| *** cmasseraf has quit IRC | 19:52 | |
| *** praefect has quit IRC | 19:53 | |
| *** cmasseraf has joined #openstack | 19:53 | |
| *** negronjl has joined #openstack | 19:54 | |
| *** Transformer has joined #openstack | 19:56 | |
| *** Transformer has quit IRC | 19:57 | |
| *** rnorwood has quit IRC | 19:58 | |
| *** stewart has quit IRC | 19:59 | |
| *** dirkx_ has quit IRC | 19:59 | |
| vidd | Kiall, any ideas? | 20:02 |
| Kiall | yea, kinda | 20:03 |
| Kiall | looks like keystone was half installing | 20:03 |
| vidd | are you part of the development team there? | 20:03 |
| Kiall | no | 20:03 |
| Kiall | wait - where is "there" ? ;) | 20:04 |
| vidd | rackspace UK | 20:04 |
| vidd | [you are from the Islands right?] | 20:04 |
| Kiall | No, I dont work for rackspace ;) | 20:05 |
| Kiall | I'm from Ireland | 20:05 |
| vidd | sorry for the UK crack then =] | 20:05 |
| Kiall | ;) | 20:05 |
| vidd | y'all dont refer to them as the "British Iles...ya call them the Invaded Lands=] | 20:08 |
| *** mattray has quit IRC | 20:10 | |
| *** anotherjesse has quit IRC | 20:10 | |
| *** anotherjesse has joined #openstack | 20:11 | |
| *** TreeStump has quit IRC | 20:13 | |
| *** duckhand has quit IRC | 20:13 | |
| *** pweiss has joined #openstack | 20:15 | |
| *** anotherjesse_ has joined #openstack | 20:15 | |
| vidd | so have you figured out what part of keystone is not installing and [more importantly]how to fix it? | 20:15 |
| *** duckhand has joined #openstack | 20:16 | |
| Kiall | vidd, waiting on the 2x images to download so I can snapshot those ;) | 20:16 |
| Kiall | PITA waiting for them everytime i revert to a previous snapshot | 20:16 |
| vidd | ah ic | 20:17 |
| vidd | well...not really...but i understand =] | 20:17 |
| *** anotherjesse has quit IRC | 20:17 | |
| *** anotherjesse_ is now known as anotherjesse | 20:17 | |
| koolhead17 | Kiall, so even on diablo quantum is needed :P | 20:17 |
| Kiall | even with that VM having at least 1Gb/s connection.. the cloud-images.ubuntu.com server is soooo sloooow | 20:17 |
| koolhead17 | Could not import dashboard.views. Error was: No module named quantum.client | 20:18 |
| Kiall | <Kiall> koolhead17, yea, dashboard can make use of quantum | 20:18 |
| Kiall | <Kiall> but does not require it | 20:18 |
| vidd | koolhead17, why not just install it? | 20:19 |
| *** anotherjesse has quit IRC | 20:19 | |
| *** dolphm_ has joined #openstack | 20:19 | |
| koolhead17 | vidd, installing it | 20:19 |
| *** rnorwood has joined #openstack | 20:23 | |
| *** dolphm has quit IRC | 20:23 | |
| Kiall | vidd, turns out i was doing something too fast .. | 20:24 |
| *** dolphm_ has quit IRC | 20:24 | |
| vidd | ? | 20:24 |
| Kiall | a sleep 5 between restarting keystone and running the data script fixed it | 20:24 |
| *** lorin1 has left #openstack | 20:24 | |
| koolhead17 | vidd, http://paste.ubuntu.com/741546/ this is why :) | 20:25 |
| Kiall | Just talking the "After Glance Complete" snapshot ;) | 20:26 |
| Kiall | vidd ^ | 20:26 |
| Kiall | based on the scripts in https://github.com/managedit/openstack-setup/tree/automated/ | 20:26 |
| Kiall | (note the branch, its not the master branch) | 20:27 |
| *** n88 has quit IRC | 20:28 | |
| cmasseraf | Hello, I'm going to start installing openstack on a 1 ethernet server... I've read it has some modifications needed, anyone can poit me which ones? (I've only installed once in a 2 ethernet PC as a test) | 20:30 |
| mjfork | cmasseraf: i am not aware of any modifications needed - what are you tryign to do | 20:31 |
| *** nati2 has joined #openstack | 20:32 | |
| *** haji has quit IRC | 20:33 | |
| *** dolphm has joined #openstack | 20:33 | |
| cmasseraf | install the cloud controller | 20:33 |
| cmasseraf | simple install | 20:34 |
| *** rsampaio has joined #openstack | 20:34 | |
| koolhead17 | vidd, you think this strange issue am having cos of virtualenv | 20:34 |
| *** dolphm has quit IRC | 20:35 | |
| mjfork | cmasseraf: then no, should be no special things for yuo to do | 20:35 |
| vidd | koolhead17, it is definantly a possibility | 20:36 |
| *** Aurelgadjo has joined #openstack | 20:37 | |
| cmasseraf | thanks everyone, will try a simple package install (since the script assumes 2 ethernets) and see what happens | 20:37 |
| *** chrism_ has joined #openstack | 20:37 | |
| juddm | folks seeing this? keystone-manage user list\n "IndexError: tuple out of range." | 20:37 |
| *** dprince has quit IRC | 20:37 | |
| chrism_ | I've got a cactus->diablo upgrade that refuses to start an instance now. They perpetually hang in "pending" status. No errors in any logs. =( | 20:38 |
| chrism_ | any ideas? | 20:38 |
| juddm | nevermind. :) | 20:39 |
| mjfork | cmasseraf: are you not installing from package? | 20:40 |
| *** Archeron has quit IRC | 20:40 | |
| *** dirkx_ has joined #openstack | 20:42 | |
| cmasseraf | the last time I installed from a script which assumed 2 ethernets | 20:51 |
| *** jeh has quit IRC | 20:51 | |
| *** jaypipes has quit IRC | 20:52 | |
| *** maplebed has joined #openstack | 20:52 | |
| maplebed | I'm setting up swift and see in the /etc/swift/*-server.conf files, it requires a bind_ip. If I leave that blank (or omit the directive) will it bind to all IPs on the host? I'm trying to deploy a stardardized config and it'd be simpler if I didn't need to modify it per-host. | 20:53 |
| *** Archeron has joined #openstack | 20:53 | |
| *** dgags has quit IRC | 20:54 | |
| vidd | maplebed, most apps on openstack bind to 0.0.0.0 | 20:55 |
| maplebed | ok. | 20:55 |
| maplebed | http://swift.openstack.org/howto_installmultinode.html keeps recommending specific IP addresses all over the place. | 20:56 |
| vidd | maplebed, i personally dont use swift =] | 20:56 |
| maplebed | gotcha. | 20:57 |
| Spirilis | maplebed: I can personally vouch that ignoring bind_ip causes it to bind 0.0.0.0 | 20:57 |
| maplebed | excellent. | 20:57 |
| maplebed | thanks! | 20:57 |
| *** katkee has quit IRC | 20:59 | |
| vidd | Kiall, looks like its working =] | 21:01 |
| vidd | what is different? | 21:01 |
| Kiall | vidd, yea, i was just about to say... | 21:01 |
| vidd | [so i can incorperate it in my scripts] | 21:01 |
| Kiall | All the scripts are working ;) | 21:01 |
| Kiall | basically - a pile of sleeps | 21:01 |
| vidd | did you add sleeps to the keystone script as well? | 21:02 |
| Kiall | all over the place ;) | 21:02 |
| vidd | what does the sleep do? is there a reason it is significant? | 21:03 |
| Kiall | yea - it lets stuff settle, DB tables get created etc etc etc | 21:05 |
| vidd | so let me guess...doing a massive keystone data dump needs to sleep after each line? | 21:06 |
| chrism_ | any ideas what would cause an instance to get stuck in "pending" state during creation on diablo? | 21:06 |
| chrism_ | no errors in any logs. =( | 21:07 |
| *** nyeates has quit IRC | 21:07 | |
| vidd | chrism_, all kinds of things can do that =] | 21:07 |
| *** jaypipes has joined #openstack | 21:07 | |
| vidd | are you on an all in one? | 21:07 |
| *** cp16net has quit IRC | 21:08 | |
| *** fulanito has joined #openstack | 21:08 | |
| chrism_ | vidd: nah, one system is api+scheduler+network and two are compute+volume | 21:09 |
| vidd | are all machines in sync time-wise? | 21:09 |
| fulanito | kiall: do you know what does Cidr stads for? | 21:09 |
| chrism_ | I believe, let me verify | 21:09 |
| vidd | time sync issues is usually the biggest culprit =] | 21:10 |
| chrism_ | yeah, within like .05 of a second | 21:10 |
| chrism_ | sorry, .005 | 21:10 |
| chrism_ | lol | 21:10 |
| fulanito | kiall: I wanto to assign a port 22 rule in the edit security group rules section in the dashboard, any advice? | 21:11 |
| vidd | so....all smiles on the services list? | 21:11 |
| chrism_ | yep | 21:11 |
| fulanito | help!!! anybody? | 21:12 |
| *** dev_sa has joined #openstack | 21:12 | |
| chrism_ | i see the query hit the compute node, and get logged in /v/l/nova/nova-compute.log, but then it just sits there | 21:12 |
| *** mgoldmann has quit IRC | 21:12 | |
| *** dev_sa has left #openstack | 21:12 | |
| vidd | Kiall, those n\ 's are suposed to be \n\ 's | 21:14 |
| fulanito | vidd: http://docs.openstack.org/diablo/openstack-compute/starter/content/Security_Overview-d1e2505.html...... in this URL do you know who is 192.168.0.1? | 21:14 |
| fulanito | vidd: sorry, its 192,168.1.1 | 21:15 |
| vidd | [space]\n\[space] | 21:15 |
| nouse | Hey guys, I'm trying to run euca-restart-instances but nova-network (Cactus release) seems to error out with this in the logs. http://pastebin.com/ii5eAdME | 21:15 |
| *** fabiand__ has left #openstack | 21:15 | |
| vidd | fulanito, thats for the private ip's | 21:17 |
| *** Turicas has quit IRC | 21:17 | |
| *** mdomsch has joined #openstack | 21:18 | |
| chrism_ | nouse: 184.75.196.201/32 is already bound to the system | 21:18 |
| nouse | chrism_: Yes, it's assigned to another instance. The instance I'm trying to restart is .200 | 21:18 |
| Kiall | vidd, I thoughts so ;) | 21:19 |
| chrism_ | hrm yeah for some reason it's trying to use the same IP again | 21:19 |
| Kiall | vidd, BTW .. I have it down to "install ubuntu, login as root, install git, clone the scripts, edit settings, run ./all-in-one.sh" ;) | 21:20 |
| *** msivanes has quit IRC | 21:22 | |
| *** haji has joined #openstack | 21:23 | |
| *** dolphm has joined #openstack | 21:23 | |
| haji | kiall I still get a route to host | 21:24 |
| haji | with ssh | 21:24 |
| Kiall | haji, this is with multiple servers, VLANs, and multi-node network right? | 21:26 |
| haji | yep | 21:26 |
| Kiall | and youre sure the switches and security group rules are right? | 21:27 |
| *** haji has quit IRC | 21:31 | |
| *** jedi4ever has quit IRC | 21:31 | |
| chrism_ | so any ideas on how to debug these VMs getting stuck in "pending" during startup? | 21:33 |
| chrism_ | I've added --debug to nova.conf and --verbose=1 i believe on all nodes | 21:33 |
| chrism_ | =( | 21:33 |
| *** nacx has joined #openstack | 21:34 | |
| *** haji has joined #openstack | 21:34 | |
| *** mdomsch has quit IRC | 21:34 | |
| haji | the vlan nova-network are working right | 21:35 |
| *** nati2 has quit IRC | 21:35 | |
| haji | the security groups rules I don't know | 21:36 |
| vidd | Kiall, mine is similar....with the settings built as you go =] | 21:37 |
| Kiall | okay, well, from the compute node, you should be able to "bypass" the rules and access the fixed_ip directly | 21:38 |
| haji | I don't which address from I should allow | 21:38 |
| Kiall | 0.0.0.0/0 | 21:39 |
| *** clopez has quit IRC | 21:40 | |
| haji | oh | 21:40 |
| *** hsmod has quit IRC | 21:40 | |
| Kiall | 0.0.0.0/0 == From anywhere... | 21:41 |
| haji | okkk | 21:41 |
| cmasseraf | when I made a nova-manage project zipfile... it returned a whron EC2_ACESS_KEY variable | 21:42 |
| cmasseraf | it returned user:project when I wanted the key | 21:44 |
| pweiss | kiall: which scripts are you using? | 21:44 |
| cmasseraf | installing by package following http://www.rhonabwy.com/wp/2011/09/23/installing-openstack-diablo-release-nova-and-glance/ | 21:44 |
| Kiall | pweiss, mine ;) | 21:45 |
| pweiss | kiall: are you sharing? | 21:45 |
| Kiall | https://github.com/managedit/openstack-setup | 21:46 |
| pweiss | thank you | 21:46 |
| cmasseraf | the problem is: I did not save that key... is there a way to retreive it? | 21:46 |
| *** Razique has joined #openstack | 21:47 | |
| *** Razique_ has joined #openstack | 21:48 | |
| *** Razique has quit IRC | 21:48 | |
| *** Razique_ is now known as Razique | 21:48 | |
| *** Razique has quit IRC | 21:49 | |
| *** nacx has quit IRC | 21:49 | |
| *** Razique has joined #openstack | 21:49 | |
| *** troytoman-away is now known as troytoman | 21:50 | |
| cmasseraf | anyway to retreive the EC2_ACCESS_KEY from it? | 21:50 |
| *** dwcramer has quit IRC | 21:52 | |
| Kiall | cmasseraf, why are you thinking nova-manage gave you the wrong thing? | 21:52 |
| cmasseraf | it gave me EC2_ACCESS_KEY as a user:project | 21:53 |
| Kiall | yea? | 21:54 |
| Razique | hi :) | 21:54 |
| Kiall | cmasseraf, have you checked if they work? or just assuming they are wrong? ;) | 21:54 |
| *** cdub has quit IRC | 21:55 | |
| cmasseraf | cloud-publish-image is giving me errors, and there was a comment on the guide I'm following about it | 21:55 |
| Kiall | I would imagine its unrelated to the EC2 key, Nova doesnt use the same format of key .. | 21:56 |
| Kiall | ..doesnt use the same format of key as EC2..* | 21:56 |
| cmasseraf | oh | 21:57 |
| cmasseraf | ok | 21:57 |
| cmasseraf | Error: Unable to run euca--describe-images. Is environment for euca- set up? | 21:57 |
| Kiall | if you type `env | grep EC2` you do you see the EC2 settings? (Just in case!) | 21:58 |
| *** rnorwood has quit IRC | 21:58 | |
| Kiall | Otherwise, run with --debug, and failing that, check the server side logs | 21:58 |
| cmasseraf | all variables are there... I'll try the --debug | 21:59 |
| *** dendrobates is now known as dendro-afk | 22:00 | |
| *** rnorwood has joined #openstack | 22:00 | |
| *** dirkx_ has quit IRC | 22:00 | |
| cmasseraf | I returned me unrecnized option --debug | 22:00 |
| Kiall | maybe its verbose, i use the nova tools rather than the eucalyptus tools myself... | 22:01 |
| vidd | cmasseraf, you have keystone installed? | 22:02 |
| *** dendro-afk is now known as dendrobates | 22:03 | |
| *** nati2 has joined #openstack | 22:03 | |
| cmasseraf | vidd: no, I don't | 22:04 |
| cmasseraf | I'm finishing the instalation of nova now | 22:04 |
| cmasseraf | Kiall: didn't work either | 22:04 |
| cmasseraf | unrecognized option | 22:04 |
| vidd | cmasseraf, have you sources your creds? | 22:04 |
| cmasseraf | yes | 22:04 |
| vidd | did you fix the EC2 env? | 22:05 |
| cmasseraf | fix? | 22:05 |
| vidd | by defaut it comes out $nova_user_name:$noav_project | 22:05 |
| cmasseraf | yes | 22:06 |
| vidd | it needs to be $REAL_KEY:$NOVA_PROJECT | 22:06 |
| cmasseraf | that's what I tought it was wrong | 22:06 |
| Kiall | really? | 22:06 |
| Kiall | I never had to change the project zipfiles | 22:06 |
| vidd | Kiall, yes...really | 22:06 |
| cmasseraf | the problem is I didn't save the real key... | 22:06 |
| Kiall | But, its been ages since I used nova without keystone | 22:06 |
| vidd | cmasseraf, you used mysql right? | 22:07 |
| cmasseraf | yes | 22:07 |
| vidd | querry your database | 22:07 |
| vidd | [or look it up in phpmyadmin] | 22:07 |
| *** zparta has quit IRC | 22:07 | |
| cmasseraf | which table? | 22:08 |
| vidd | use i believe | 22:08 |
| vidd | give me a sec to verify | 22:08 |
| vidd | yes...users | 22:09 |
| cmasseraf | access_key? | 22:09 |
| vidd | yep | 22:09 |
| *** deshantm_laptop has quit IRC | 22:10 | |
| cmasseraf | changed it and got the same error | 22:11 |
| vidd | you need to re-sourse your creds | 22:12 |
| cmasseraf | did it | 22:12 |
| *** nati2 has quit IRC | 22:12 | |
| vidd | what error message? | 22:12 |
| cmasseraf | Unable to run euca--describe-images. Is environment for euca- set up? | 22:12 |
| vidd | then your creds are not properly sources | 22:12 |
| vidd | did you copy the *.pem 's to your ~/ folder? | 22:13 |
| cmasseraf | ~/nova/creds/ | 22:13 |
| vidd | run this: | 22:14 |
| *** clauden_ has quit IRC | 22:14 | |
| vidd | cp ~/nova/creds/*.pem ~/ | 22:14 |
| Kiall | vidd, wait. what | 22:14 |
| Kiall | no | 22:14 |
| *** clauden_ has joined #openstack | 22:14 | |
| vidd | Kiall, what do you mean "no"? | 22:14 |
| Kiall | look at the novarc file, the first thing it does is figure out what folder its in so it can set the ENV vars to the correct paths.. | 22:15 |
| *** haji has quit IRC | 22:15 | |
| vidd | if you look at the ENV it says the PEM is in the ~/folder | 22:15 |
| Kiall | they, without a shadow of a doubt, do not need to be in the ~ folder | 22:15 |
| cmasseraf | #NOVA_KEY_DIR is setted correctly | 22:15 |
| vidd | Kiall, i always has to move them =\ | 22:16 |
| Kiall | Before I started with keystone, I always unzipped nova.zip to ~/.nova .. and never had to edit the file.. | 22:16 |
| vidd | cmasseraf, humor me | 22:16 |
| cmasseraf | I coppied anyway... no change | 22:17 |
| *** stewart has joined #openstack | 22:17 | |
| cmasseraf | but the variable is setted correctly | 22:17 |
| vidd | cmasseraf, ok...delete the files that were copied (or if you moved them move them back) | 22:17 |
| vidd | some variable is not correct =\ | 22:18 |
| cmasseraf | I'll tray extracting it again | 22:18 |
| cmasseraf | try* | 22:18 |
| vidd | cmasseraf, are you typing"--" in that command? | 22:19 |
| *** marrusl has quit IRC | 22:20 | |
| cmasseraf | I'm running "cloud-publish-tarball $image imageBucket" | 22:20 |
| vidd | cmasseraf, is that failing? | 22:20 |
| cmasseraf | yes | 22:20 |
| vidd | what is that error message? | 22:21 |
| cmasseraf | because euca-describe-images fails | 22:21 |
| cmasseraf | the one I've sent | 22:21 |
| cmasseraf | Unable to run euca--describe-images. Is environment for euca- set up? | 22:21 |
| *** mrevell has joined #openstack | 22:21 | |
| *** clauden_ has quit IRC | 22:21 | |
| vidd | cloud-publish-tarball should not be trying to do a describe image | 22:22 |
| vidd | and "euca--describe-images" is an invalid command | 22:22 |
| WormMan | there's a reason I avoid the cloud publish and euca tarball stuff | 22:23 |
| Kiall | WormMan++ | 22:24 |
| Kiall | why use tools for eucalyptus with openstack, rather than openstacks own tools | 22:24 |
| fulanito | anybody knows how to make an ssh to an instance? | 22:28 |
| fulanito | the ip of the instance does not respond | 22:28 |
| Kiall | fulanito, have you setup the security rules to allow ssh? | 22:31 |
| fulanito | yep, i did it for 0.0.0.0/0 | 22:31 |
| Kiall | and you're sure the instance has booted correctly? | 22:32 |
| Kiall | (you checked the console logs..) | 22:32 |
| *** haji has joined #openstack | 22:32 | |
| haji | nope i still can't | 22:32 |
| Kiall | haji, your internet connection must suck ;) constant ping timeouts! | 22:32 |
| *** ldlework has quit IRC | 22:32 | |
| haji | i can ping my vlan and br100 interface between nodes | 22:33 |
| haji | i can ping the public interface between nodes | 22:33 |
| fulanito | yep is running and it ask for a user and password, | 22:33 |
| haji | but i cant ping the fixed address from the instance | 22:33 |
| Kiall | fulanito, what do you mean it asks for a user/pass? in VNC? | 22:33 |
| haji | nor the floating ip | 22:34 |
| fulanito | yes | 22:34 |
| Kiall | fulanito, ok .. thats normal so ;) | 22:34 |
| Kiall | fulanito, is it one of the stock ubuntu images? or something else? | 22:34 |
| fulanito | stock ubuntu images | 22:34 |
| cmasseraf | NOVA_API_KEY should be userName or RealKey? | 22:35 |
| Kiall | cmasseraf, its been a while since i used nova without keystone, so I cant remember.. But I do remember not having to edit the novarc file | 22:35 |
| Kiall | fulanito, try enabling ping aswell and see if you can ping it | 22:36 |
| Kiall | haji, honestly, I dont know.. | 22:36 |
| *** mikeyp is now known as mikeyp-lurking | 22:36 | |
| Kiall | haji, brctl show has the instance and vlan listed as bridged to br100? | 22:37 |
| Kiall | haji, also.. eth1 is your public? and eth0 your VLAN right? | 22:37 |
| haji | eth1 is my public int | 22:39 |
| haji | yep | 22:39 |
| fulanito | kiall: I have already enable port 22 for 0.0.0.0/0 ip | 22:39 |
| Kiall | Maybe there is something up with that, I've only noticed a setting for the vlan interface | 22:39 |
| haji | br100 8000.02163e2981d no vlan100 | 22:40 |
| Kiall | fulanito, sure, but maybe something is wrong with SSH on the instance | 22:40 |
| Kiall | try enabling ICMP -1:-1 from 0.0.0.0/0 and pinging it | 22:40 |
| fulanito | when I ping it on a compute node it says de destination host is unreachable | 22:40 |
| Kiall | Ah .. the compute node should be able to ping the fixed IP of the node, regardless of security group rules | 22:41 |
| Kiall | haji, do you have vlan_interface=eth0 and public_interface=eth1 set in your nova.conf's? | 22:42 |
| haji | yeah | 22:42 |
| haji | i tried your updated scripts | 22:43 |
| fulanito | it still does not respond | 22:43 |
| Kiall | haji, really? I setup a 2x nic VM a few mins ago and they work a charm | 22:45 |
| Kiall | they dont have a public_interface setting though | 22:45 |
| Kiall | (I just added it now that I found it) | 22:45 |
| haji | oh | 22:46 |
| haji | let me add it | 22:46 |
| haji | then | 22:46 |
| *** mahmoh has joined #openstack | 22:46 | |
| cmasseraf | just droped the database and restart the configuration to see if I did anything wrong | 22:49 |
| *** mrevell has quit IRC | 22:49 | |
| haji | the ip address on the vlan interface should 172.16.X.X something | 22:51 |
| haji | cuz i have 192.168.X.X | 22:51 |
| haji | :o | 22:51 |
| Kiall | haji, working? | 22:53 |
| *** redconnection has joined #openstack | 22:53 | |
| haji | nop | 22:54 |
| Kiall | http://paste.openstack.org/show/3409/ | 22:54 |
| Kiall | thats my /etc/network/interfaces | 22:54 |
| Kiall | swap the eth0's obv | 22:54 |
| Kiall | but ignore the vlan interface IP | 22:54 |
| *** rsampaio has quit IRC | 22:55 | |
| Kiall | set it to something completely unused, something that will never be used. | 22:55 |
| Kiall | I use 172 for fixed IP stuff, and "real" IPs for Floating IPs and physical servers | 22:55 |
| Kiall | hence - the 192 address thats never seen a single packet of traffic | 22:56 |
| *** nati2 has joined #openstack | 22:56 | |
| *** rsampaio has joined #openstack | 22:58 | |
| *** robbiew has quit IRC | 22:58 | |
| cmasseraf | now my nova-manage is trhowing an exection on FlagsError(e) | 22:59 |
| cmasseraf | I'll do a complete reinstall and use keystone | 22:59 |
| Kiall | haji, http://paste.openstack.org/show/3410/ | 23:00 |
| Kiall | cmasseraf, may I suggest the easy route? ;) https://github.com/managedit/openstack-setup/ | 23:00 |
| ohnoimdead | are there some magical instructions somewhere on how to acquire and upload a oneiric image to openstack with glance and keystone (and using the glance cli tool not euca2ools)? | 23:01 |
| *** GheRivero_ has quit IRC | 23:02 | |
| ejat | ohnoimdead : i just pass that phase .. by studying backward :) | 23:02 |
| *** aliguori has quit IRC | 23:02 | |
| Kiall | glance add -A KEYSTONE_TOKEN name="bla" is_public=true disk_format=qcow2 container_format=ovf < filename | 23:02 |
| * ejat wish to have GUI for it rather than the glance CLI tools | 23:02 | |
| Kiall | from memory | 23:02 |
| pweiss | kiall…for your scripts….anything other than 11.10 and 2 nics? will it run in a VM? | 23:03 |
| ejat | Kiall :) | 23:03 |
| Kiall | pweiss, everything bar the compute node will.. | 23:03 |
| ohnoimdead | yeah, adding image upload is on the roadmap for horizon. but right now, i'm pulling out what little hair i have left. | 23:03 |
| Kiall | and .. i suppose, that could probably work too | 23:03 |
| cmasseraf | Kiall: scripts for each instalation? are they customizable? | 23:03 |
| pweiss | thanks | 23:03 |
| ohnoimdead | Kiall: is filename the sec .img? | 23:03 |
| haji | kiall: http://pastebin.com/xQfgG9DL | 23:04 |
| Kiall | yea | 23:04 |
| Kiall | cmasseraf, semi... it makes some assumptions around VLAN network rather than FlatDHCP, and 2x nics etc | 23:04 |
| ejat | ohnoimdead : noted | 23:04 |
| Kiall | cmasseraf, all the settings are here https://github.com/managedit/openstack-setup/blob/master/settings | 23:04 |
| cmasseraf | I use only one ethenet device here | 23:04 |
| Kiall | (ie - not much..) | 23:04 |
| *** stewart has quit IRC | 23:04 | |
| Kiall | But it gives you a working nova+keystone+glace+dashboard at the end ;) | 23:05 |
| cmasseraf | =) | 23:05 |
| cmasseraf | I'm having a look | 23:05 |
| Kiall | Installed from .deb's rather than git clones and random crap spewed accross /ust/local ;) | 23:05 |
| *** jj0hns0n has joined #openstack | 23:06 | |
| Kiall | haji, 192.168.3.1 is the public facing IP of the conttroller node? | 23:06 |
| haji | i dont have the floating ips on the my public interface | 23:06 |
| haji | no thats the private interface | 23:06 |
| cmasseraf | the only problem is that my server only has eth0 | 23:07 |
| haji | or vlan interface connected to the switch | 23:07 |
| Kiall | k | 23:07 |
| Kiall | so that means 10.10.10.2 is the public IP of the controller node? | 23:07 |
| cmasseraf | so how would I setup your script? | 23:07 |
| haji | yeah | 23:07 |
| ejat | if using the devstack … how do i start all the component after restart the machine ? | 23:08 |
| Kiall | cmasseraf, 100% clean install (really - do this step), install git, clone the repo, edit settings.. run the files in the order the readme says | 23:08 |
| ejat | without rerun the script again ? | 23:08 |
| *** imsplitbit has quit IRC | 23:09 | |
| Kiall | haji, honestly, I dont know! | 23:09 |
| Kiall | Maybe take a step back and try with a single node first ;) | 23:09 |
| haji | :o | 23:10 |
| Kiall | I just managed a full setup in 40mins with those scripts (excluding OS install).. its not that long ;) | 23:10 |
| Kiall | and 10 mins of that was downloading lucid and oneiric images ;) | 23:10 |
| ejat | Kiall :) | 23:10 |
| pweiss | kiall…was that an all in one? | 23:11 |
| haji | fml | 23:11 |
| Kiall | Yea | 23:11 |
| Kiall | give the experimental "all-in-one.sh" a go ;) | 23:11 |
| Kiall | literally - edit settings | 23:11 |
| Kiall | hit go | 23:11 |
| Kiall | done | 23:11 |
| ejat | same goes to me .. | 23:11 |
| haji | i already tried devstack on a vm | 23:11 |
| pweiss | im working on it…wanting for 11.10 to finish downloading | 23:11 |
| Kiall | haji, sure.. what I mean is.. | 23:12 |
| cmasseraf | kiall: what I'm saying is | 23:12 |
| Kiall | start with an all in one | 23:12 |
| haji | i used this to create floating ips: sudo nova-manage floating create --ip_range=10.10.10.224/27 | 23:12 |
| Kiall | then add a second compute node | 23:12 |
| haji | aaah | 23:12 |
| haji | ooks | 23:12 |
| Kiall | then remove compute from the controller if you want | 23:12 |
| cmasseraf | I only have eth0 what i'd use here: PUBLIC_INTERFACE="eth0" # Interface for Floating IP traffic | 23:12 |
| cmasseraf | VLAN_INTERFACE="eth1" # Interface for Fixed IP VLAN traffic | 23:12 |
| cmasseraf | I only have the floating trafic | 23:12 |
| Kiall | cmasseraf, Honestly, I've never tried a single interface setup.. I simply have no clue if setting them both to eth0 will work | 23:12 |
| Kiall | maybe it will.. but i have no idea! | 23:13 |
| ohnoimdead | Kiall: i can upload the oneiric .img into glance and it shows up fine. but when i try to launch an instance based on that image it has "0 bytes" ram and 0 GB disk. is there some metadata i'm missing or something? | 23:13 |
| ejat | Kiall : i did with the single interface | 23:13 |
| pweiss | kiall…what does your topology look like? 2 nics…one switch or two? | 23:13 |
| mjfork | ohnoimdead: i am goign to guess a cached base image exists already with no content | 23:14 |
| Kiall | pweiss, the simple version of Public SW -> eth0 Server eth1 -> VLAN Switch | 23:14 |
| Kiall | s/of/is/ | 23:14 |
| cmasseraf | ejat: How did you do it? | 23:14 |
| Kiall | pweiss, but in real life, its 2x switches and nested VLANs ;) | 23:15 |
| pweiss | makes sense | 23:15 |
| *** fulanito has quit IRC | 23:15 | |
| ejat | cmasseraf : http://paste.ubuntu.com/741728/ | 23:16 |
| ejat | i edit the stack.sh as per instruction | 23:16 |
| mjfork | ohnoimdead: look for a file in the images directory that has the ID of the base image that will not start | 23:16 |
| ejat | but u still can change the fixed_range | 23:16 |
| ejat | the u associate to the instance | 23:17 |
| ohnoimdead | mjfork: where does glance store images? | 23:18 |
| mjfork | not glance, but on the nova node | 23:18 |
| mjfork | ohnoimdead: basically, nova caches the image at first use. | 23:18 |
| mjfork | the images directory is on the same level as instances | 23:19 |
| ohnoimdead | mjfork: ok, same question: where does nova cache the image? | 23:20 |
| *** robix has joined #openstack | 23:20 | |
| mjfork | on my redhat box, /var/lib/libvirt/ | 23:21 |
| robix | Seeing this error in Horizon in keypairs and other places. Using devstack: unexpected keyword argument 'api_key' | 23:21 |
| ejat | mjfork under qemu right ? | 23:22 |
| mjfork | ejat: what do you mean? | 23:22 |
| cloudfly | hrmm | 23:22 |
| ejat | /var/lib/libvirt/qemu ? | 23:23 |
| cloudfly | what's the clean way to restart a devstack install after a reboot? | 23:23 |
| *** TheOsprey has quit IRC | 23:23 | |
| ohnoimdead | robix: that is due to https://bugs.launchpad.net/horizon/+bug/891442 you can work around by updating python-novaclient in the horizon venv (tools/with_venv.sh pip install python-novaclient --update) | 23:23 |
| *** Shentonfreude has quit IRC | 23:24 | |
| ohnoimdead | robix: in horizon/openstack-dashboard/ | 23:24 |
| ejat | cloudfly : rerun .. it will drop the db then will recreate back if im not mistaken .. | 23:25 |
| mjfork | ejat: yes, that could be. whereever the instances are created. you can tell from the logs too. | 23:25 |
| mikeyp-lurking | cloudfly: I hacked an if/else into the script, to just restart the services and screen sessions ear the end of the script. | 23:26 |
| ejat | ok noted | 23:26 |
| ejat | $ ./with_venv.sh pip install python-novaclient --update | 23:26 |
| ejat | ./with_venv.sh: line 4: ./../.dashboard-venv/bin/activate: No such file or directory | 23:26 |
| *** mikeyp-lurking is now known as mikeyp | 23:26 | |
| *** colon_D has joined #openstack | 23:27 | |
| mjfork | ejat: ohnoimdead the file you need to delete serves as the base of the qcow2 disks | 23:27 |
| ohnoimdead | ejat: from one layer up: cd .. then tools/with_venv.sh blah blah | 23:27 |
| ohnoimdead | mjfork: ok i'm looking at the libvirt.xml and this thing is totally borked. it's not even close to something functional. whatever part of nova that is building the libvirt.xml is expecting information that is not being provided... | 23:28 |
| mikeyp | cloudfly: a rerun also works, but it rebuilds the db, not sure if it has other effects if you've tweaked things. | 23:28 |
| mjfork | got an example? | 23:29 |
| ejat | .dashboard-venv | 23:29 |
| ejat | ?? | 23:29 |
| ohnoimdead | mjfork: https://gist.github.com/1374933 | 23:30 |
| mjfork | ohnoimdead: looks god to me | 23:30 |
| *** rsampaio has quit IRC | 23:30 | |
| *** rsampaio has joined #openstack | 23:31 | |
| mjfork | ohnoimdead: "cd /opt/stack/nova/nova/..//instances/instance-00000002/disk" | 23:31 |
| ohnoimdead | mjfork: shouldn't there be a disk node? | 23:31 |
| mjfork | i see 2 of them? | 23:31 |
| ohnoimdead | mjfork: good lord, nm not looking at the right thing… | 23:32 |
| *** redconnection has quit IRC | 23:32 | |
| ohnoimdead | mjfork: still, no love on firing up the instance | 23:32 |
| *** dgags has joined #openstack | 23:33 | |
| *** ahasenack has quit IRC | 23:34 | |
| ohnoimdead | ejat: run "tools/with_venv.sh pip install python-novaclient --upgrade" from the horizon/openstack-dashboard directory | 23:34 |
| mjfork | ohnoimdead: do you have any running instances? | 23:34 |
| ohnoimdead | well nova says this one is "active" | 23:35 |
| ejat | stack@thinkpad:/opt/stack/horizon/openstack-dashboard$ tools/with_venv.sh pip install python-novaclient --upgrade | 23:35 |
| ejat | tools/with_venv.sh: line 4: tools/../.dashboard-venv/bin/activate: No such file or directory | 23:35 |
| mjfork | ohnoimdead: can you terminate it? | 23:35 |
| ohnoimdead | ejat: do you have the virtual env installed? is there a .dashboard-venv directory in openstack-dashboard? if not go all the way up to the horizon root and run "./run_tests.sh" | 23:36 |
| ohnoimdead | mjfork: i could before (this is the fourth or fifth try) but i'll try again | 23:36 |
| ejat | nope | 23:36 |
| *** llang629 has joined #openstack | 23:36 | |
| mjfork | ohnoimdead: once it is terminated, see if images or instances directories have any contents | 23:37 |
| *** llang629 has left #openstack | 23:37 | |
| ohnoimdead | mjfork: terminated | 23:37 |
| mjfork | ohnoimdead: is there anything in instances or images directories? | 23:38 |
| ohnoimdead | mjfork: here's something odd, i only have an instances directory, not an images… instances now contains "_base" | 23:39 |
| mjfork | ahh, yes. what is in _base | 23:40 |
| mjfork | sorry, was doing this from memory and was confusing _base with imagse :( | 23:40 |
| ohnoimdead | mjfork: stuff | 23:40 |
| *** anotherjesse has joined #openstack | 23:40 | |
| mjfork | delete it | 23:40 |
| ohnoimdead | mjfork: https://gist.github.com/1374967 | 23:41 |
| mjfork | do an ls -l on it, i bet it has 0 length | 23:41 |
| ohnoimdead | mjfork: including the _base dir or just its contents? | 23:41 |
| mjfork | just the contents is fine | 23:42 |
| ohnoimdead | mjfork: https://gist.github.com/1374971 | 23:42 |
| ohnoimdead | mjfork: still want me to delete it? | 23:42 |
| andrewbogott | Does this message mean that my instances have to have access to the outside internet in order to properly start up? | 23:42 |
| andrewbogott | waiting for metadata service at http://169.254.169.254/2009-04-04/meta-data/instance-id | 23:42 |
| mjfork | ohnoimdead: yes, i would. even though it lokos good. | 23:43 |
| andrewbogott | Or is 169.254.169.254 supposed to map to my host, somehow? | 23:43 |
| mjfork | ohnoimdead: gievs you a fresh start. be sure to turn on verbose logging in nova.conf on this create | 23:43 |
| andrewbogott | Hm... nm, may've found an answer... | 23:44 |
| *** dubsquared has quit IRC | 23:45 | |
| cloudfly | i'll give it a shot | 23:45 |
| cloudfly | thanks | 23:45 |
| andrewbogott | well, wait, this thread says that that shouldn't happen with flat dhcp (which I am using) | 23:46 |
| mjfork | andrewbogott: no, that is a speical url for resolving metadata that (i believe) has an iptables entry to resolve | 23:47 |
| *** redconnection has joined #openstack | 23:48 | |
| andrewbogott | Yeah, I see. So something is messed up with the network on my host. (Foolishly, I expected devstack to take care of that for me.) | 23:49 |
| mjfork | single machine sinstall? | 23:49 |
| *** rsampaio has quit IRC | 23:49 | |
| ejat | ohnoimdead : Successfully installed python-novaclient httplib2 argparse prettytable | 23:50 |
| ejat | thanks | 23:50 |
| *** rnorwood1 has joined #openstack | 23:51 | |
| *** AlanClark has quit IRC | 23:51 | |
| *** rsampaio has joined #openstack | 23:51 | |
| ohnoimdead | ejat: cool, glad that helped | 23:51 |
| ohnoimdead | mjfork: same deal but here's all of the logs i could get which may, or may not, be useful: https://gist.github.com/1375001 | 23:52 |
| *** nerens has quit IRC | 23:54 | |
| mjfork | one waning, but it looks ok. "ignoring error injecting data into image 4 (Mapped device was not found (we can only inject raw disk images): /dev/mapper/nbd15p1)" | 23:54 |
| *** rnorwood has quit IRC | 23:54 | |
| *** Razique has quit IRC | 23:54 | |
| andrewbogott | mjfork: Yep, single machine, two nics. | 23:54 |
| mjfork | ohnoimdead: what status does nova report for you instance? and what does virsh list show. | 23:54 |
| mjfork | andrewbogott: i would have expected it to just work. is nova-network running? | 23:55 |
| andrewbogott | yep, network is running. Nothing obvious in the logfile, but I'm still looking... | 23:56 |
| *** troytoman is now known as troytoman-away | 23:56 | |
| *** ejat has quit IRC | 23:56 | |
| mjfork | andrewbogott: run "ip addr show" is 169.254.169.254 shown? | 23:58 |
| *** ejat has joined #openstack | 23:58 | |
| mjfork | for me it is on lo | 23:58 |
| andrewbogott | Nope, lo is 127.0.0.1. | 23:59 |
| *** daMaestro has joined #openstack | 23:59 | |
| daMaestro | is it possible to do http://docs.amazonwebservices.com/AmazonS3/latest/dev/index.html?HTTPPOSTForms.html with openstack swift3? | 23:59 |
| daMaestro | it looks like it should be | 23:59 |
| andrewbogott | mjfork: I do have HOST_IP=192.168.56.20 set because I wanted horizon to run in a place where I could reach it. Maybe that's confounding the devstack setup. | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!