| *** davlap has joined #openstack-dev | 00:02 | |
| *** RatTehReaL has joined #openstack-dev | 00:02 | |
| *** RatTehReaL has left #openstack-dev | 00:02 | |
| *** davlap is now known as davlaps | 00:02 | |
| *** paulormg has quit IRC | 00:02 | |
| *** littleidea has quit IRC | 00:04 | |
| *** maplebed has quit IRC | 00:05 | |
| *** littleidea has joined #openstack-dev | 00:12 | |
| *** littleidea has quit IRC | 00:16 | |
| *** blakehaggerty has joined #openstack-dev | 00:17 | |
| *** littleidea has joined #openstack-dev | 00:18 | |
| *** bhall has quit IRC | 00:19 | |
| notmyname | justinsb: ping | 00:24 |
|---|---|---|
| *** jakedahn_zz is now known as jakedahn | 00:27 | |
| *** reed has quit IRC | 00:28 | |
| *** johnpostlethwait has quit IRC | 00:29 | |
| *** markvoelker has quit IRC | 00:35 | |
| *** dolphm has joined #openstack-dev | 00:40 | |
| notmyname | justinsb: ah, nm. looks like chmouel gave you all the info you need | 00:40 |
| *** mszilagyi has quit IRC | 00:47 | |
| *** cp16net has joined #openstack-dev | 00:48 | |
| *** cp16net has quit IRC | 00:49 | |
| ayoung | termie, heckj, vishy Ryan_Lane I have a blueprint for PKI support. http://wiki.openstack.org/PKI but I am not sure it is publically visible yet | 00:52 |
| Ryan_Lane | ah, interesting | 00:52 |
| *** hattwick has quit IRC | 00:55 | |
| ayoung | Ryan_Lane, can you see it> | 00:55 |
| ayoung | > is ? from my fat fingers | 00:55 |
| Ryan_Lane | yeah. can see it | 00:56 |
| ayoung | Ryan_Lane, OK, I'll create a BLueprint on Launchpad and link to it. | 00:57 |
| *** hattwick has joined #openstack-dev | 00:59 | |
| heckj | ayoung: perfect! Link that puppy up! | 00:59 |
| ayoung | https://blueprints.launchpad.net/keystone/+spec/pki | 00:59 |
| ayoung | heckj, there ya go | 01:00 |
| ayoung | heckj, a word about Dogtag | 01:00 |
| ayoung | It is a full CA as I said | 01:00 |
| ayoung | and iti getting an appropriate REST API | 01:00 |
| ayoung | It is the Artist formerly known as Netscape Certificate server, it is really old, solid code, with all that implies | 01:00 |
| heckj | interesting! | 01:01 |
| heckj | It'll be a great topic at the design summit! | 01:01 |
| ayoung | I'd suspect that we would use it to get things going, and at some point, people are going to complain about it being in Java | 01:01 |
| ayoung | at that point, assuming we are using it, we can propose a 'G' version that is a subset rewrite using the same techns as the rest of openstack | 01:01 |
| ayoung | something kind of along the line of what Rackspace did with the internal Keystone, only here it would be an open source project and part of daily development | 01:02 |
| *** jdurgin has quit IRC | 01:02 | |
| ayoung | heckj, Oh, I should add a link to the REST api ...will do so now | 01:03 |
| ayoung | heckj, Ryan_Lane http://pki.fedoraproject.org/wiki/Dogtag_Future_Directions | 01:06 |
| Ryan_Lane | dogtag is *very* dependent on fedora | 01:07 |
| ayoung | Ryan_Lane, not really | 01:08 |
| ayoung | the current build system is | 01:08 |
| ayoung | but it is a Java enterprise App | 01:09 |
| Ryan_Lane | I've looked at installing it on ubuntu before, and it didn't look like it would be easy | 01:09 |
| ayoung | with some reworking, it should be pretty agnostic of the underlying distribution | 01:09 |
| ayoung | It will take wokr | 01:09 |
| ayoung | work | 01:09 |
| * Ryan_Lane nods | 01:09 | |
| ayoung | I'd like to replace the CMake build system with Maven | 01:09 |
| Ryan_Lane | (I'd love to have dogtag on ubuntu, I've wanted to use it for a while) | 01:09 |
| ayoung | that will pull in the Distro specific build process | 01:09 |
| ayoung | I am assuming that Debian is Maven friendly | 01:10 |
| *** villep has quit IRC | 01:10 | |
| *** hattwick has quit IRC | 01:11 | |
| *** martine has joined #openstack-dev | 01:12 | |
| heckj | ayoung: it is, we've been building out PDF based docs with it | 01:13 |
| *** littleidea has quit IRC | 01:13 | |
| ayoung | heckj, so the Fedora bindings are to support certain deployment requirements for certain customers, and a lot of that is not necessary for most users | 01:17 |
| *** pixelbeat has quit IRC | 01:17 | |
| ayoung | We're going to go with the approach that Dogtag requires tomcat, and gets installed as an app using the default tomcat instance | 01:18 |
| *** hattwick has joined #openstack-dev | 01:19 | |
| *** martine has quit IRC | 01:20 | |
| williamherry | mysql gone away bug show on my node | 01:22 |
| *** hattwick has quit IRC | 01:23 | |
| williamherry | i find this bug reported at 2011-1-22, still not fixed ? | 01:23 |
| *** hattwick has joined #openstack-dev | 01:25 | |
| *** dalang has quit IRC | 01:26 | |
| *** rods has quit IRC | 01:27 | |
| *** ncode has joined #openstack-dev | 01:27 | |
| *** ncode has joined #openstack-dev | 01:27 | |
| *** Ryan_Lane has quit IRC | 01:28 | |
| *** ayoung has quit IRC | 01:28 | |
| *** ayoung has joined #openstack-dev | 01:29 | |
| *** deshantm has quit IRC | 01:29 | |
| *** hattwick has quit IRC | 01:30 | |
| *** dolphm has quit IRC | 01:31 | |
| *** Mandell has quit IRC | 01:32 | |
| *** hattwick has joined #openstack-dev | 01:32 | |
| *** PotHix has quit IRC | 01:32 | |
| *** negronjl has quit IRC | 01:34 | |
| *** vincentricci has left #openstack-dev | 01:35 | |
| *** crobinso has quit IRC | 01:36 | |
| justinsb | On Debian, /sbin isn't in the path, so mkfs isn't found with devstack. Should I fix in devstack (by adding /sbin to PATH) or in nova (by changing mkfs to /sbin/mkfs)? | 01:42 |
| *** blakehaggerty has quit IRC | 01:44 | |
| *** dolphm has joined #openstack-dev | 01:45 | |
| *** shevek_ has quit IRC | 01:51 | |
| dtroyer | justinsb: is this on precise? if so, it's a change in sudo that caused this; chmouel is working on it. Our workaround is adding this to shudders (or something in shudders.d): | 01:51 |
| dtroyer | Defaults secure_path="${HOME}/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin" | 01:51 |
| dtroyer | s/shudders/sudoers/ ^^^ autcorrect! | 01:52 |
| dtroyer | justinsb: duh, you said Debian…same issue though maybe? | 01:53 |
| *** cp16net has joined #openstack-dev | 02:02 | |
| *** heckj has quit IRC | 02:08 | |
| *** bencherian has quit IRC | 02:09 | |
| *** negronjl has joined #openstack-dev | 02:12 | |
| *** villep has joined #openstack-dev | 02:16 | |
| *** negronjl has quit IRC | 02:16 | |
| *** negronjl has joined #openstack-dev | 02:17 | |
| *** mnewby has quit IRC | 02:21 | |
| *** novas0x2a|laptop has quit IRC | 02:21 | |
| *** adjohn has quit IRC | 02:22 | |
| *** danwent has quit IRC | 02:32 | |
| *** mattstep has quit IRC | 02:39 | |
| *** mattstep has joined #openstack-dev | 02:40 | |
| *** milner has quit IRC | 02:46 | |
| med_ | does Launch Instance in devstack still have an off by one error (if more than one requested)? | 02:48 |
| * med_ guesses that's a horizon question | 02:48 | |
| *** roge has joined #openstack-dev | 02:48 | |
| *** littleidea has joined #openstack-dev | 02:54 | |
| *** milner has joined #openstack-dev | 03:00 | |
| justinsb | dtroyer: Actually, I don't believe the command is run as root. It is mkfs in nova/virt/libvirt/utils.py. | 03:03 |
| *** ywu has quit IRC | 03:09 | |
| *** agonella has quit IRC | 03:24 | |
| *** paulstallworth has joined #openstack-dev | 03:25 | |
| *** ncode has quit IRC | 03:25 | |
| *** Ryan_Lane has joined #openstack-dev | 03:32 | |
| *** dolphm has quit IRC | 03:38 | |
| *** dolphm has joined #openstack-dev | 03:39 | |
| *** Mandell has joined #openstack-dev | 03:46 | |
| *** dubsquared has quit IRC | 03:57 | |
| *** roge has quit IRC | 03:59 | |
| *** dolphm has quit IRC | 04:03 | |
| *** adjohn has joined #openstack-dev | 04:20 | |
| *** mnewby has joined #openstack-dev | 04:21 | |
| *** spiffxp has quit IRC | 04:24 | |
| *** danwent has joined #openstack-dev | 04:30 | |
| *** dolphm has joined #openstack-dev | 04:32 | |
| *** mattstep has quit IRC | 04:35 | |
| *** mattstep has joined #openstack-dev | 04:35 | |
| *** mattstep has quit IRC | 04:48 | |
| *** mattstep has joined #openstack-dev | 04:49 | |
| *** cp16net has quit IRC | 04:52 | |
| *** cp16net has joined #openstack-dev | 04:53 | |
| *** littleidea has quit IRC | 05:07 | |
| *** Vek has quit IRC | 05:13 | |
| *** novas0x2a|laptop has joined #openstack-dev | 05:20 | |
| *** littleidea has joined #openstack-dev | 05:23 | |
| *** mpe- has joined #openstack-dev | 05:24 | |
| *** hattwick has quit IRC | 05:32 | |
| *** dolphm has quit IRC | 05:33 | |
| *** milner has quit IRC | 05:36 | |
| *** eglynn_ has quit IRC | 05:37 | |
| *** eglynn_ has joined #openstack-dev | 05:37 | |
| *** danwent has quit IRC | 05:39 | |
| *** milner has joined #openstack-dev | 05:49 | |
| *** gyee has quit IRC | 05:50 | |
| *** eglynn__ has joined #openstack-dev | 05:52 | |
| *** shevek_ has joined #openstack-dev | 05:54 | |
| *** eglynn_ has quit IRC | 05:54 | |
| anotherjesse | mtaylor: idea - what about not letting merges of reviews with WIP in title land? | 05:55 |
| *** novas0x2a|laptop has quit IRC | 05:56 | |
| *** cp16net has left #openstack-dev | 05:59 | |
| *** novas0x2a|laptop has joined #openstack-dev | 05:59 | |
| *** jakedahn is now known as jakedahn_zz | 06:02 | |
| *** Mandell has quit IRC | 06:10 | |
| *** cp16net has joined #openstack-dev | 06:21 | |
| *** davlaps has quit IRC | 06:22 | |
| *** mnewby has quit IRC | 06:22 | |
| *** cp16net has quit IRC | 06:24 | |
| mtaylor | anotherjesse: interesting | 06:40 |
| anotherjesse | would make commiting a change to openstack-ci to implement WIP complicated | 06:41 |
| mtaylor | anotherjesse: also - there is a feature in gerrit 2.2.2 (which we are testing for deployment, there's an internal mysql regression blocking deployment right now) | 06:41 |
| anotherjesse | ;p | 06:41 |
| mtaylor | anotherjesse: which allows you to upload a 'draft' change | 06:41 |
| anotherjesse | mtaylor: nice - I know a several folks use the review process for WIP | 06:41 |
| mtaylor | anotherjesse: and then when you're ready you can move it from draft to real status | 06:41 |
| mtaylor | anotherjesse: yeah, and I'd like to support that | 06:41 |
| anotherjesse | my team likes it - before we were doing pull requests for team stuff, then sending a single review | 06:41 |
| mtaylor | anotherjesse: one of the guys at hp is working on a jenkins job for an internal gerrit that would watch for two reviews with +2 and then do the approval +1 vote ... I wonder if an easy short-term solution would be to do a similar jenkins job to register a -2 blocking vote | 06:43 |
| mtaylor | when it sees WIP | 06:43 |
| mtaylor | anotherjesse: also - I'm coming down to SF week after next to chat with Adrian Cole about the jclouds jenkins plugin (first version works now, btw) | 06:43 |
| mtaylor | anotherjesse: I should swing by and chat with you too | 06:44 |
| anotherjesse | mtaylor: that deserves a blog post and beers! | 06:44 |
| mtaylor | anotherjesse: yes, as soon as I have tested it myself, I am going to blog the SHIT out of it | 06:44 |
| anotherjesse | mtaylor: definitely - rax has new offices - will pm you the address | 06:44 |
| mtaylor | anotherjesse: sweet | 06:44 |
| *** adjohn has quit IRC | 06:45 | |
| mtaylor | anotherjesse: https://bugs.launchpad.net/openstack-ci/+bug/954754 | 06:50 |
| uvirtbot | Launchpad bug 954754 in openstack-ci "block merges marked WIP" [Wishlist,Triaged] | 06:50 |
| *** paulstallworth has quit IRC | 06:50 | |
| *** cp16net_ has joined #openstack-dev | 06:50 | |
| anotherjesse | mtaylor: depending on timeline for 2.2.2, it isn't worth doing - if it is < 1 month I would just do the upgrade | 06:51 |
| *** adjohn has joined #openstack-dev | 06:52 | |
| *** cp16net_ has quit IRC | 06:53 | |
| *** adjohn has quit IRC | 06:56 | |
| *** adjohn has joined #openstack-dev | 06:59 | |
| *** adjohn has quit IRC | 07:00 | |
| *** belmoreira has joined #openstack-dev | 07:03 | |
| mtaylor | anotherjesse: I REALLY want it to be < 1 month ... we're running on MyISAM until we get 2.2.2 installed :( | 07:06 |
| mtaylor | anotherjesse: but there is a CRAZY sql regression that only really is a problem for installations which are large like ours ... the groups page loads EVERY user for every group with no caching (which for us is many many users * many many groups) | 07:08 |
| mtaylor | anotherjesse: but we'll try to get it done | 07:08 |
| anotherjesse | surprised android doesn't have issues with it then | 07:09 |
| mtaylor | ah - android has jumped to 2.3 already | 07:12 |
| *** Vek has joined #openstack-dev | 07:15 | |
| *** justinsb has quit IRC | 07:24 | |
| *** justinsb has joined #openstack-dev | 07:30 | |
| *** Vek has quit IRC | 07:35 | |
| *** anotherjesse is now known as anotherjesse_zz | 07:46 | |
| *** milner has quit IRC | 07:51 | |
| *** hattwick has joined #openstack-dev | 07:59 | |
| *** littleidea has quit IRC | 07:59 | |
| *** novas0x2a|laptop has quit IRC | 08:00 | |
| *** berendt has joined #openstack-dev | 08:04 | |
| *** milner has joined #openstack-dev | 08:04 | |
| *** eglynn has joined #openstack-dev | 08:08 | |
| *** eglynn__ has quit IRC | 08:10 | |
| *** eglynn_ has joined #openstack-dev | 08:13 | |
| *** reidrac has joined #openstack-dev | 08:14 | |
| *** eglynn has quit IRC | 08:16 | |
| *** phorgh has joined #openstack-dev | 08:34 | |
| *** novas0x2a|laptop has joined #openstack-dev | 08:47 | |
| *** derekh has joined #openstack-dev | 08:55 | |
| *** eglynn__ has joined #openstack-dev | 08:58 | |
| *** bencherian has joined #openstack-dev | 08:59 | |
| *** kbringard has joined #openstack-dev | 08:59 | |
| *** eglynn_ has quit IRC | 09:01 | |
| *** zigo has joined #openstack-dev | 09:09 | |
| LarsErikP | anyone know why this occur? diablo 2011.3.1 http://paste.openstack.org/show/9254/ | 09:16 |
| *** Ryan_Lane has quit IRC | 09:22 | |
| *** kbringard has quit IRC | 09:31 | |
| *** oneiroi has joined #openstack-dev | 09:32 | |
| *** Vek has joined #openstack-dev | 09:32 | |
| *** mpe- has quit IRC | 09:37 | |
| *** LinuxJedi has quit IRC | 09:39 | |
| *** LinuxJedi has joined #openstack-dev | 09:41 | |
| williamherry | do not see this before, but did you notice that it says | 09:43 |
| williamherry | Instance '605' failed to spawn. Is virtualization enabled in the BIOS? Details | 09:43 |
| LarsErikP | williamherry: it is enabled. It worked yesterday... | 09:45 |
| *** eglynn has joined #openstack-dev | 09:48 | |
| justinsb | LarsErikP: It looks more like a problem with there not being any ips in the array. I think the BIOS message is just a 'best guess' whenever something goes wrong. | 09:48 |
| LarsErikP | justinsb: that was what i thought to.. but i have no instances running at all, so the fixed_ip range should be all clear... I've run 100's of instances on this stack before... | 09:50 |
| *** eglynn__ has quit IRC | 09:50 | |
| justinsb | LarsErikP: I don't really know. Looks like the code is still in trunk, so no obvious bugfixes. I did spot this bug: | 09:55 |
| justinsb | Bug #953708 | 09:55 |
| uvirtbot | Launchpad bug 953708 in nova "nova w/ FlatDHCP isn't fully deallocating fixed_ips" [High,In progress] https://launchpad.net/bugs/953708 | 09:55 |
| LarsErikP | i'm running VLan manager.. | 09:56 |
| justinsb | LarsErikP: According to that bug, if you run 'nova-manage fixed list' you can check that your IPs aren't exhausted | 09:56 |
| LarsErikP | that command ouptputs this http://paste.openstack.org/show/9264/ | 09:56 |
| LarsErikP | not sure why there are only three left.... as i said. Each tenant is configured to get a /24 net. And as I said, no instances are running atm | 09:57 |
| justinsb | LarsErikP: It is odd that there are 3, I think. I ran it on mine, and it listed hundreds (well, 255 I think) | 09:58 |
| justinsb | LarsErikP: But I really don't know VLAN manager | 09:58 |
| justinsb | LarsErikP: And I promised to stay away from networking for a while, as I tend to break things | 09:58 |
| zykes- | Kiall: around for a general ubuntu packaging question? | 09:59 |
| justinsb | LarsErikP: Maybe check the nova-network log for any error messages? | 09:59 |
| *** novas0x2a|laptop has quit IRC | 10:00 | |
| justinsb | LarsErikP: Ah - I think you are out of addresses. Most managers reserve 2, VLAN reserves an extra one for the VPN server. | 10:02 |
| justinsb | LarsErikP: Now, as for why you're out of addresses, there I'm definitely out of my depth | 10:02 |
| LarsErikP | justinsb: yeah, I run select * from fixed_ips where instance_id is not null\G | 10:03 |
| LarsErikP | it listed every single ip in the 172.16.0.0/24 net... | 10:03 |
| justinsb | LarsErikP: So they're assigned to instances that have been shutdown? | 10:03 |
| LarsErikP | they're assigned to instances that have been destroyed | 10:04 |
| LarsErikP | with 'nova delete <id>' | 10:04 |
| justinsb | LarsErikP: Sounds like a bug to me. I'd file a bug report and reference 953708. | 10:08 |
| *** eglynn has quit IRC | 10:10 | |
| *** eglynn has joined #openstack-dev | 10:11 | |
| *** maploin has joined #openstack-dev | 10:11 | |
| *** maploin has quit IRC | 10:11 | |
| *** maploin has joined #openstack-dev | 10:11 | |
| justinsb | LarsErikP: It could also be related to bug #907898 | 10:13 |
| uvirtbot | Launchpad bug 907898 in nova "You can't specify target table 'fixed_ips' for update" [High,Fix released] https://launchpad.net/bugs/907898 | 10:13 |
| *** devananda has quit IRC | 10:17 | |
| *** pixelbeat has joined #openstack-dev | 10:23 | |
| LarsErikP | think i walked around it... ran update fixed_ips set instance_id=NULL where instance_id < 1000; | 10:23 |
| LarsErikP | and voilla | 10:23 |
| justinsb | LarsErikP: Sounds good. When you next shutdown a machine, I would check the nova network log to see if there's an error involving freeing the fixed_ip | 10:25 |
| LarsErikP | i think i misconfigured the --force_dhcp_release. In diablo it need =True. my config lacked that | 10:25 |
| *** tryggvil has joined #openstack-dev | 10:25 | |
| *** n0ano has quit IRC | 10:25 | |
| *** dneary has joined #openstack-dev | 10:34 | |
| *** dneary has quit IRC | 10:34 | |
| *** dneary has joined #openstack-dev | 10:34 | |
| *** milner has quit IRC | 10:48 | |
| *** maploin has quit IRC | 10:55 | |
| *** paulormg has joined #openstack-dev | 11:01 | |
| *** milner has joined #openstack-dev | 11:01 | |
| *** novas0x2a|laptop has joined #openstack-dev | 11:08 | |
| *** eglynn_ has joined #openstack-dev | 11:09 | |
| *** eglynn has quit IRC | 11:09 | |
| *** markvoelker has joined #openstack-dev | 11:16 | |
| *** davlaps has joined #openstack-dev | 11:21 | |
| LarsErikP | Since Vlan-manager creates a new vlan for each tenant.. That will be an issue if you have more tenants than possible VLAN-id's? | 11:28 |
| *** dneary has quit IRC | 11:30 | |
| *** rods has joined #openstack-dev | 11:32 | |
| *** bepernoot has joined #openstack-dev | 11:34 | |
| zykes- | LarsErikP: possible vlan id's being ? | 11:37 |
| LarsErikP | there is a maximum of 4096 vlans? am i not right? | 11:38 |
| LarsErikP | minus the reserved ones... | 11:38 |
| zykes- | 1-4095 yeah | 11:38 |
| zykes- | but you also have the new stuff "VXLAN" | 11:38 |
| LarsErikP | so, what if you hace 5000 tenants? | 11:38 |
| LarsErikP | *have | 11:38 |
| LarsErikP | is that something openstack supports? | 11:39 |
| zykes- | LarsErikP: checkout http://blogs.cisco.com/openatcisco/integrating-vxlan-in-openstack-quantum/ | 11:39 |
| zykes- | http://blogs.cisco.com/datacenter/digging-deeper-into-vxlan/ | 11:39 |
| LarsErikP | nice. thanks! | 11:39 |
| zykes- | and maybe Nicira are doing some stuff as well ? | 11:40 |
| zykes- | + HP | 11:40 |
| *** bsza has joined #openstack-dev | 11:42 | |
| *** maploin has joined #openstack-dev | 11:49 | |
| *** maploin has quit IRC | 11:49 | |
| *** maploin has joined #openstack-dev | 11:49 | |
| *** ghe_ has quit IRC | 11:53 | |
| *** davidkranz has joined #openstack-dev | 11:59 | |
| *** berendt has quit IRC | 12:03 | |
| *** berendt has joined #openstack-dev | 12:04 | |
| *** lts has joined #openstack-dev | 12:17 | |
| *** martine has joined #openstack-dev | 12:17 | |
| *** dprince has joined #openstack-dev | 12:24 | |
| *** littleidea has joined #openstack-dev | 12:32 | |
| *** littleidea has quit IRC | 12:33 | |
| *** eglynn__ has joined #openstack-dev | 12:48 | |
| *** eglynn_ has quit IRC | 12:50 | |
| *** RobertLaptop has joined #openstack-dev | 12:54 | |
| *** novas0x2a|laptop has quit IRC | 12:59 | |
| *** Glacee has joined #openstack-dev | 13:02 | |
| *** RobertLaptop has left #openstack-dev | 13:02 | |
| *** Glace_ has joined #openstack-dev | 13:04 | |
| *** Glacee has quit IRC | 13:06 | |
| *** littleidea has joined #openstack-dev | 13:12 | |
| *** Mandell has joined #openstack-dev | 13:22 | |
| *** j05h has quit IRC | 13:23 | |
| *** andrewsmedina has joined #openstack-dev | 13:26 | |
| *** cmagina has quit IRC | 13:28 | |
| *** cmagina has joined #openstack-dev | 13:29 | |
| *** Mandell has quit IRC | 13:31 | |
| *** deshantm has joined #openstack-dev | 13:31 | |
| *** crobinso has joined #openstack-dev | 13:40 | |
| soren | Daviey: Have you guys thought about how to handle database upgrades at all? | 13:40 |
| *** dneary has joined #openstack-dev | 13:46 | |
| *** dneary has quit IRC | 13:46 | |
| *** dneary has joined #openstack-dev | 13:46 | |
| *** andrewsmedina has quit IRC | 13:47 | |
| *** joesavak has joined #openstack-dev | 13:55 | |
| *** andrewsmedina has joined #openstack-dev | 13:59 | |
| *** roge has joined #openstack-dev | 14:01 | |
| *** dolphm has joined #openstack-dev | 14:01 | |
| annegentle | ttx: around? | 14:08 |
| *** novas0x2a|laptop has joined #openstack-dev | 14:09 | |
| *** roge has quit IRC | 14:14 | |
| *** roge has joined #openstack-dev | 14:15 | |
| *** dolphm has quit IRC | 14:17 | |
| Daviey | soren: How so? | 14:29 |
| Daviey | soren: nova does migrations on startup, right? | 14:29 |
| Daviey | soren: But yes, the initial database migration blueprint was mine.. so i did do some thought :) | 14:30 |
| *** crobinso has quit IRC | 14:30 | |
| *** dtroyer is now known as dtroyer_zzz | 14:31 | |
| *** kbringard has joined #openstack-dev | 14:32 | |
| *** reed has joined #openstack-dev | 14:32 | |
| *** zzed has joined #openstack-dev | 14:37 | |
| *** tryggvil has quit IRC | 14:43 | |
| *** dtroyer_zzz is now known as dtroyer | 14:45 | |
| ttx | annegentle: yes | 14:49 |
| annegentle | ttx: now that the Summit site is open, shall I send out a call for OpenStack Today submissions? | 14:49 |
| ttx | annegentle: to submit them outside of the site ? Or inside the site ? | 14:51 |
| annegentle | ttx: I think outside the site, right? Or inside? | 14:51 |
| ttx | annegentle: outside is ok. | 14:51 |
| annegentle | ttx: I was thinking a simple ether pad for signup | 14:51 |
| annegentle | ttx: and I'll start email requests today | 14:51 |
| ttx | annegentle: sounds good | 14:51 |
| annegentle | ttx: just looking at timing, getting another's viewpoint :) thanks. | 14:52 |
| *** gyee has joined #openstack-dev | 14:54 | |
| *** hhoover has joined #openstack-dev | 14:55 | |
| *** hhoover has left #openstack-dev | 14:56 | |
| kbringard | I seem to recall work being done awhile back to have images converted to raw when they get cached on the compute node and instances brought up… did that ever come to be? | 14:58 |
| *** ghe_ has joined #openstack-dev | 15:03 | |
| *** dolphm has joined #openstack-dev | 15:04 | |
| *** eglynn__ has quit IRC | 15:08 | |
| *** eglynn__ has joined #openstack-dev | 15:08 | |
| *** Hitesh__ has joined #openstack-dev | 15:12 | |
| *** maplebed has joined #openstack-dev | 15:18 | |
| vishy | chmouel: make sure tenant_to_connect only works if reseller_admin role is set | 15:19 |
| vishy | i propped my part | 15:20 |
| vishy | chmouel: https://review.openstack.org/5338 | 15:20 |
| *** eglynn__ has quit IRC | 15:21 | |
| *** eglynn__ has joined #openstack-dev | 15:21 | |
| chmouel | vishy: well I can't really make it this is when passed to swift_auth | 15:21 |
| chmouel | vishy: you can do the same thing with curl | 15:21 |
| chmouel | vishy: if youg et a username/token and connect to another AUTH_account | 15:22 |
| chmouel | vishy: (you get) | 15:22 |
| vishy | does that mean that any user can do this? | 15:23 |
| vishy | or it only works with reseller admin token? | 15:23 |
| chmouel | only works with reseller admin | 15:23 |
| chmouel | normal user would get rejected if it does not have reseller right | 15:23 |
| chmouel | I was actually just cleaning this up before sending it | 15:24 |
| *** kbringard1 has joined #openstack-dev | 15:25 | |
| *** kbringard has quit IRC | 15:29 | |
| *** Hitesh__ has quit IRC | 15:30 | |
| *** hitesh__ has joined #openstack-dev | 15:33 | |
| *** kbringard1 has quit IRC | 15:33 | |
| *** deshantm_ has joined #openstack-dev | 15:36 | |
| *** kbringard has joined #openstack-dev | 15:36 | |
| *** vincentricci has joined #openstack-dev | 15:37 | |
| *** deshantm_ has quit IRC | 15:38 | |
| *** danwent has joined #openstack-dev | 15:38 | |
| *** deshantm has quit IRC | 15:40 | |
| vishy | chmouel: sorry got disconnected for a minute | 15:41 |
| ttx | vishy: what do you think about bug 954923 ? | 15:42 |
| uvirtbot | Launchpad bug 954923 in nova "In essex-4: bad link to users table into nova database when using euca-xxx commands" [Undecided,New] https://launchpad.net/bugs/954923 | 15:42 |
| andrewbogott | bcwaldon: If you are not waist-deep in last-minute essex bugs, I'd appreciate your comments on the API in change 5292. | 15:43 |
| vishy | ttx: that seems incorrect | 15:44 |
| vishy | ttx: might be true with noauth | 15:44 |
| vishy | let me look | 15:44 |
| *** AlanClark has joined #openstack-dev | 15:44 | |
| *** AlanClark has quit IRC | 15:44 | |
| vishy | ttx: correct, only true with noauth | 15:45 |
| notmyname | Daviey: ping. want to talk about swift releases? | 15:45 |
| chmouel | vishy: this is the big comment i have passed in there http://pastie.org/private/iqoqcog6knqtawgn9ufl9a | 15:46 |
| vishy | chmouel: great | 15:46 |
| vishy | chmouel: the comment is quite long! | 15:46 |
| vishy | if you propose it I will ping heckj about getting it in | 15:47 |
| vishy | ttx: commented on the bug | 15:50 |
| *** n0ano has joined #openstack-dev | 15:50 | |
| *** andrewsmedina has quit IRC | 15:51 | |
| dtroyer | notmyname: could you have a look at https://review.openstack.org/5320 for the CLI stuff we proposed yesterday? | 15:53 |
| chmouel | vishy: cool i am creating some scripts to help people to get started if they wanted to test it for review | 15:54 |
| vishy | chmouel: using devstack? | 15:54 |
| chmouel | not devstack yet but i guess i should :) | 15:55 |
| *** rkukura has quit IRC | 15:57 | |
| *** davlaps has quit IRC | 15:58 | |
| *** novas0x2a|laptop has quit IRC | 15:59 | |
| *** tryggvil_ has joined #openstack-dev | 15:59 | |
| Daviey | notmyname: can you give me an hr? | 15:59 |
| notmyname | Daviey: sure (or maybe a bit later--I'll probably be at lunch in an hour) | 16:00 |
| *** zzed has quit IRC | 16:01 | |
| *** deshantm has joined #openstack-dev | 16:01 | |
| bcwaldon | andrewbogott: I'll check it out | 16:01 |
| andrewbogott | thanks! | 16:01 |
| bcwaldon | andrewbogott: I'll do a more thorough review, but I do have one piece of immediate feedback | 16:03 |
| *** markvoelker has quit IRC | 16:03 | |
| bcwaldon | andrewbogott: nova/api/openstack/compute/contrib is solely for api extensions | 16:03 |
| bcwaldon | feel free to add a nova/sharedfs directory or something for your drivers/manager | 16:03 |
| andrewbogott | So I should stash the drivers and support code elsewhere? | 16:03 |
| bcwaldon | andrewbogott: yep | 16:04 |
| andrewbogott | OK, hang on... doesn't 'extension' imply 'optional component'? And shouldn't all the code that's in support of an optional component be in a location that indicates that? | 16:04 |
| andrewbogott | e.g. contrib/sharedfs? | 16:04 |
| bcwaldon | andrewbogott: I don't think contrib is necessary, is it? | 16:05 |
| bcwaldon | andrewbogott: this is going to have to wait for folsom, btw | 16:05 |
| andrewbogott | bcwaldon: Are we going to try to keep a distinction between 'core nova' files, and 'nova extension' files? | 16:06 |
| andrewbogott | bcwaldon: I agree, re: folsom. | 16:06 |
| bcwaldon | andrewbogott: maybe vishy can comment here, but we don't really make a distinction | 16:07 |
| bcwaldon | andrewbogott: we use the 'contrib' folder for extensions, but thats limited to the context of the core api spec, not nova | 16:07 |
| andrewbogott | :) I agree that we don't currently make a distinction, but I think we should. Mainly because I want to have the ability to contribute 'marginal' code without having to get everyone on board. That seems like exactly what an 'extension' should be. | 16:08 |
| *** sannes has joined #openstack-dev | 16:08 | |
| vishy | andrewbogott: we need to have an extensive discussion at the summit for how to handle this | 16:09 |
| andrewbogott | I understand the distinction between code and api. But I still think it may be useful to have a distinction in code as well. | 16:09 |
| bcwaldon | andrewbogott: api extensions that live in the core codebase are intended for future inclusion in the next versions of the compute api | 16:09 |
| andrewbogott | vishy: Yeah, fair enough :) | 16:09 |
| vishy | andrewbogott: I'd like code like that to remain separate from core code if possible | 16:09 |
| bcwaldon | andrewbogott: would you prefer to see the solidfire iscsi driver in a contrib folder? | 16:09 |
| vishy | chmouel: does https://review.openstack.org/#change,4696 work with your change and my change? | 16:10 |
| vishy | if so, that is probably plenty to test. | 16:11 |
| andrewbogott | bcwaldon: I'm not sure. Maybe. It is sort of weird that all those drivers are jammed together in one file. | 16:11 |
| bcwaldon | andrewbogott: okay, so I think this is a bigger discussion to be had at the summit | 16:11 |
| andrewbogott | I imagine a world where, if I want to add a custom driver or extension, I just drop in those files in the appropriate place and add a commandline flag and it goes. If I have to edit python code, that seems awkward. | 16:11 |
| chmouel | vishy: yeah it does I use that, it's not great for handling error case (ie: wrong access key/port etc..) | 16:11 |
| andrewbogott | But, yeah, no need to solve this problem here and now. I'll be sure to show up for that discussion when it happens. | 16:11 |
| andrewbogott | And, really, moving the files into a 'shardfs' directory as you suggest would essentially accomplish what I'm saying anyway :) | 16:12 |
| *** eglynn__ has quit IRC | 16:12 | |
| bcwaldon | andrewbogott: really, that code shouldnt live with nova if its not owned by the community, right? | 16:13 |
| bcwaldon | andrewbogott: we dont nova to be a repository for possibly-working code | 16:13 |
| bcwaldon | andrewbogott: ...like vsa | 16:13 |
| vishy | chmouel: if you can ship a localrc that makes that work on first run i think we're good | 16:13 |
| andrewbogott | bcwaldon: There might be room for a middle ground between 'third party' and 'nova core'. That's what I was imagining at least. | 16:14 |
| bcwaldon | dprince: sorry to override you on that merge prop | 16:15 |
| andrewbogott | Certainly my hope is that the sharedfs stuff will be owned by nova; that just depends on >1 person is interested in it though. | 16:15 |
| bcwaldon | dprince: need a comment in the code explaining what it's doing | 16:15 |
| dprince | bcwaldon: fine by me. | 16:15 |
| bcwaldon | andrewbogott: ok, would you be able to run a summit discussion on it? | 16:15 |
| andrewbogott | bcwaldon: On sharedfs, or on extensibility? | 16:16 |
| bcwaldon | andrewbogott: sharedfs, specifically | 16:16 |
| bcwaldon | andrewbogott: I think the extensibility discussion is already planned | 16:16 |
| andrewbogott | Sure. | 16:16 |
| chmouel | vishy: well I still neeed to create a user with reseller right and use that in novarc (for devstack) | 16:16 |
| bcwaldon | even if it's just what, why, how | 16:16 |
| dprince | bcwaldon: this is a show stopper for my testing... (had to steer the battleship around it)... but I suppose another hour or so won't hurt. | 16:16 |
| andrewbogott | bcwaldon: That is, if it needs to be promoted :) | 16:16 |
| bcwaldon | dprince: if you understand the code, you can add a comment | 16:16 |
| bcwaldon | dprince: I personally just need help understanding it | 16:16 |
| bcwaldon | dprince: doesnt look like eglynn is around | 16:17 |
| vishy | chmouel: seems like that should be added to the admin user by default in keystone_data | 16:17 |
| andrewbogott | bcwaldon: Either way, I'm interested in your opinion about whether or not you think my design is stupid :) | 16:17 |
| dprince | bcwaldon: Eoghan is adding a comment now I believe. Stay tuned... | 16:17 |
| chmouel | vishy: not conditional if swift is isntalled? | 16:17 |
| bcwaldon | dprince: ok, great | 16:17 |
| bcwaldon | andrewbogott: I'll do a full review for you, but it'll probably have to wait for a bit today due to the folsom-target | 16:18 |
| andrewbogott | bcwaldon: At your leisure. I'm on to other things atm anyway. Thanks again. | 16:18 |
| bcwaldon | andrewbogott: no problem | 16:18 |
| *** mnewby has joined #openstack-dev | 16:20 | |
| chmouel | vishy: https://review.openstack.org/5340 | 16:21 |
| *** heckj has joined #openstack-dev | 16:21 | |
| *** Mandell has joined #openstack-dev | 16:23 | |
| *** berendt has quit IRC | 16:24 | |
| bcwaldon | dprince: approved | 16:24 |
| *** roge has quit IRC | 16:26 | |
| dprince | bcwaldon: thanks. Glad we got a comment in there too. | 16:27 |
| bcwaldon | dprince: it's hard to be that guy, but I had to do it :) | 16:28 |
| vishy | heckj: this would be great to target: https://bugs.launchpad.net/nova/+bug/954505 | 16:31 |
| ttx | vishy: been reviewing a few New bugs, added some to essex-rc-potential list for further investigation | 16:31 |
| uvirtbot | Launchpad bug 954505 in nova "Nova s3 should be able to talk to swift " [Medium,In progress] | 16:31 |
| heckj | vishy: word | 16:32 |
| *** anotherjesse_zz is now known as anotherjesse | 16:32 | |
| *** reidrac has quit IRC | 16:33 | |
| *** ayoung has quit IRC | 16:33 | |
| vishy | heckj: needed along with a change in nova for ec2 register -> nova -> swift | 16:33 |
| vishy | chmouel: it is just an extra role right? | 16:34 |
| vishy | chmouel: we add example roles in keystone_data.sh so adding it always with a note seems fine to me | 16:34 |
| heckj | bcwaldon: when you have a few, would love your eyes on https://review.openstack.org/#change,5322 | 16:34 |
| bcwaldon | heckj: F5 | 16:34 |
| heckj | bcwaldon: Thanks! | 16:35 |
| vishy | heh he beat you | 16:35 |
| heckj | why am I not surprised :-) | 16:37 |
| *** spiffxp has joined #openstack-dev | 16:37 | |
| heckj | vishy: is the role setup needed already in devstack? I'm doc' n today, so I want to describe that in the setup docs. | 16:37 |
| vishy | heckj: chmouel can answer that | 16:38 |
| vishy | i think it needs a reseller admin role | 16:38 |
| vishy | i asked chmouel to add it to his patch here: https://review.openstack.org/#change,4696 | 16:38 |
| *** bepernoot has quit IRC | 16:38 | |
| heckj | Ah- ringing a bell, I remember chmouel's review related to that | 16:38 |
| annegentle | heckj: yeah I could swear I saw a set of "roles" someone was configuring and I can't find it now. | 16:41 |
| *** oneiroi has quit IRC | 16:42 | |
| *** cp16net_ has joined #openstack-dev | 16:43 | |
| *** oneiroi has joined #openstack-dev | 16:43 | |
| *** ghe_ has quit IRC | 16:44 | |
| *** oneiroi has quit IRC | 16:44 | |
| heckj | bwaldon, anotherjesse: https://review.openstack.org/#change,5322 updated | 16:44 |
| *** ayoung has joined #openstack-dev | 16:48 | |
| *** ghe_ has joined #openstack-dev | 16:49 | |
| *** maplebed has quit IRC | 16:52 | |
| *** jdurgin has joined #openstack-dev | 16:52 | |
| *** mszilagyi has joined #openstack-dev | 16:54 | |
| chmouel | heckj, vishy: yes that's correct just an extra role | 16:54 |
| chmouel | heckj, vishy: ResellerAdmin | 16:55 |
| *** maplebed has joined #openstack-dev | 16:55 | |
| jaypipes | heckj: you had some questions for me that I missed in the meeting yesterday evening? | 16:55 |
| jaypipes | heckj: about tempest? | 16:55 |
| *** eglynn has joined #openstack-dev | 16:56 | |
| *** PotHix has joined #openstack-dev | 16:56 | |
| *** dalang has joined #openstack-dev | 16:56 | |
| heckj | yeah - was just curious where things were re: status on the project. I haven't paid any direct attention, but have been looking forward to it being more online. | 16:56 |
| *** devananda has joined #openstack-dev | 16:57 | |
| Kiall | zykes-: I am now.. | 16:57 |
| eglynn | bcwaldon, dprince: seems random 404s from the swift instance used by the jenkins are now holding up the merge of #5333 | 16:57 |
| eglynn | https://jenkins.openstack.org/job/gate-glance-python26/85/console | 16:57 |
| eglynn | different test failed in https://jenkins.openstack.org/job/gate-glance-python26/83/console, again with 404 | 16:58 |
| bcwaldon | eglynn: yeah, I rekicked the job hoping it was transient | 16:58 |
| bcwaldon | eglynn: might have to let it sit for a few then retrigger | 16:58 |
| jaypipes | heckj: trying to get a gate job set up for it today, working with jeblair on that this afternoon... won't necessarily cause it to gate trunks, but should be able to trigger it from trunk merges and start seeing some output. | 16:58 |
| eglynn | bcwaldon: k | 16:58 |
| jaypipes | heckj: takes around 20 minutes to run it, so it's not a quick little job ;) | 16:58 |
| heckj | jaypipes: that would be fantastic! | 16:58 |
| *** andrewsmedina has joined #openstack-dev | 16:58 | |
| jaypipes | heckj: yeah! long time comin! | 16:58 |
| bcwaldon | eglynn: actually, it just went through | 16:58 |
| bcwaldon | eglynn: well, the job reported success | 16:58 |
| *** tryggvil_ has quit IRC | 16:59 | |
| *** tryggvil__ has joined #openstack-dev | 16:59 | |
| eglynn | bcwaldon: cool! | 16:59 |
| eglynn | bcwaldon: we'll have to get on top of those swift failures tho' ... o'wise obviously we'll end up with a lot of merge delays | 16:59 |
| bcwaldon | eglynn: I'll keep poking at it | 16:59 |
| *** maploin` has joined #openstack-dev | 17:02 | |
| *** dolphm has quit IRC | 17:02 | |
| *** maploin has quit IRC | 17:02 | |
| *** dneary has quit IRC | 17:09 | |
| *** jeblair has quit IRC | 17:12 | |
| *** mnewby has quit IRC | 17:13 | |
| rmk | Noticed there are newer dnsmasq versions in the nova PPAs -- is there a reason for that? | 17:14 |
| *** cp16net_ has quit IRC | 17:14 | |
| *** dolphm has joined #openstack-dev | 17:15 | |
| *** maploin` has quit IRC | 17:17 | |
| vishy | justinsb: ping | 17:17 |
| jaypipes | davidkranz: ping | 17:17 |
| Ash | jaypipes: ping | 17:18 |
| jaypipes | Ash: pong | 17:18 |
| justinsb | vishy: Morning! | 17:18 |
| Ash | oh good, jaypipes is responding to ping | 17:19 |
| jaypipes | Ash: what can I do for you? | 17:19 |
| * Ash closes out alert ticket | 17:19 | |
| Ash | oh, nothing. i was just amused by all the pings. don't mind me ;) | 17:19 |
| vishy | justinsb: so the libvirt issue you are running into | 17:20 |
| vishy | why does it only break on flat? | 17:20 |
| jaypipes | Ash: :) | 17:20 |
| Ash | :D | 17:20 |
| justinsb | vishy: The 0xfe one, right? I'm not really sure why it breaks - it's deep into networking mysteries | 17:20 |
| vishy | justinsb: well it doesn't seem to break anyone else | 17:21 |
| vishy | so i'm hesitant to revert it without understanding why | 17:21 |
| justinsb | vishy: I think maybe on non-FlatManager we run something which responds to ARP locally? | 17:21 |
| *** bepernoot has joined #openstack-dev | 17:22 | |
| justinsb | vishy: Do we have an integration test which runs vs FlatManager? | 17:22 |
| vishy | justinsb: the vms should respond to arp | 17:22 |
| vishy | justinsb: no, so the issue is they can't arp? | 17:22 |
| justinsb | vishy: The VMs ask for the ARP of the gateway, the ARP request gets sent onto the network, the response comes back, but the guest never sees the response | 17:23 |
| *** rkukura has joined #openstack-dev | 17:23 | |
| vishy | are you sure this is a mac address issue? | 17:23 |
| justinsb | vishy: It could be that the gateway on other networking modes is the host machine? | 17:23 |
| vishy | justinsb: well we have versions without | 17:23 |
| *** shevek_ has quit IRC | 17:23 | |
| vishy | justinsb: but maybe no one has tried it | 17:23 |
| justinsb | vishy: Not sure. But when I reverted the patch (first byte -> 0x02), or changed the first byte to 0xfc, it worked again immediately | 17:24 |
| vishy | justinsb: flat_dhcp passing dhcp_option=3,<ip> gives it another gateway | 17:24 |
| *** novas0x2a|laptop has joined #openstack-dev | 17:24 | |
| vishy | so should we switch to 0xfc ? | 17:24 |
| justinsb | vishy: I'd be happy with a switch to 0xfc | 17:24 |
| justinsb | vishy: It's high, it avoids libvirt's 0xfe, and it 'works for me' with FlatManager | 17:24 |
| vishy | justinsb: although it is really odd that it doesn't work with 0xfe | 17:25 |
| *** cp16net_ has joined #openstack-dev | 17:25 | |
| vishy | can you send something to ml about it to see if we should report it upstream? | 17:25 |
| justinsb | vishy: I think it's that dmesg error | 17:25 |
| justinsb | vishy: Sure - good idea! | 17:25 |
| justinsb | vishy: Hopefully someone that really understands networking will be able to answer | 17:26 |
| vishy | justinsb: 0xfc has the proper locally admin and unicast bits set yes? | 17:26 |
| justinsb | vishy: I think so, though I didn't even know about them until you posted that comment. Networking black magic | 17:26 |
| vishy | justinsb: incorrect | 17:28 |
| vishy | we need to use fa | 17:28 |
| justinsb | vishy: Doh! You're right. Haven't had enough coffee yet! | 17:29 |
| *** bepernoot has quit IRC | 17:34 | |
| vishy | justinsb: here is the workaround patch: https://review.openstack.org/5351 | 17:35 |
| vishy | if you want to refer to it in the email | 17:35 |
| vishy | I will also have our networking guy look at it. | 17:36 |
| *** dolphm has quit IRC | 17:37 | |
| vishy | justinsb: I also think it is strange that libvirt is setting mac if one is specified | 17:39 |
| justinsb | vishy: I think that libvirt is helping us out, and avoiding the bug. It may be that this bug was never a bug, though then the linked bug is doubly problematic. | 17:40 |
| vishy | justinsb: if it is actually overriding the one you specify that seems like a bug | 17:40 |
| justinsb | vishy: I don't know which MAC address actually makes it onto the wire | 17:40 |
| justinsb | vishy: If it's 0xfe, and not the one we specify, I could imagine that could cause us problems if we're not aware of it | 17:41 |
| justinsb | vishy: So I think I sent us on a wild goose chase on the MAC address with the bridge connectivity - sorry! | 17:41 |
| justinsb | vishy: But at least I understand networking a little better (maybe) | 17:41 |
| vishy | no one ever understands networking | 17:41 |
| vishy | :) | 17:41 |
| justinsb | vishy: I hope somebody understands it. Maybe we'll flush out the experts with the ML question! | 17:42 |
| *** eglynn_ has joined #openstack-dev | 17:42 | |
| justinsb | vishy: Some good news - I checked the ARP tables, and it looks like the MAC address on the wire is the MAC address of the guest, not the bridge | 17:44 |
| vishy | justinsb: so it is just tricking the bridge | 17:45 |
| justinsb | vishy: Although that of course begs the question of what the MAC address of the vnetX device is used for | 17:45 |
| *** eglynn has quit IRC | 17:45 | |
| vishy | justinsb: weird | 17:45 |
| *** rgoodwin has quit IRC | 17:45 | |
| justinsb | vishy: I'm just not going to touch networking for a while. I'm self-imposing a few weeks in the networking penalty box. | 17:45 |
| *** stuntmachine has joined #openstack-dev | 17:46 | |
| anotherjesse | vishy: https://review.openstack.org/#change,5345 for folsom? | 17:48 |
| vishy | anotherjesse: already commented | 17:49 |
| anotherjesse | vishy: you commented that the flag should be default until folsom, I was wondering if the patch itself should be folsom | 17:49 |
| *** adjohn has joined #openstack-dev | 17:52 | |
| *** rgoodwin has joined #openstack-dev | 17:56 | |
| *** ayoung has quit IRC | 17:57 | |
| *** devananda has quit IRC | 17:57 | |
| *** devananda has joined #openstack-dev | 17:57 | |
| zykes- | Kiall: now ? :) | 17:58 |
| Kiall | Yah | 17:58 |
| zykes- | nvm, i figured it out :) | 17:58 |
| Kiall | lol | 17:58 |
| zykes- | it was some upstart thing earlier | 17:58 |
| zykes- | :o | 17:58 |
| *** johnpostlethwait has joined #openstack-dev | 18:02 | |
| *** johnpostlethwa-1 has joined #openstack-dev | 18:04 | |
| *** johnpostlethwait has quit IRC | 18:04 | |
| *** dprince has quit IRC | 18:04 | |
| *** jakedahn_zz is now known as jakedahn | 18:12 | |
| chmouel | vishy: I have sent a mail to the ml about the testing with those patches and it seems to fail around glance | 18:14 |
| *** jakedahn is now known as jakedahn_zz | 18:15 | |
| vishy | chmouel: glance is failing now? | 18:15 |
| chmouel | vishy: not sure tbh, see my output | 18:15 |
| chmouel | s/output/paste/ | 18:15 |
| chmouel | http://pastie.org/private/10pta1v5txfzwjjzfehukg | 18:15 |
| chmouel | vishy: i have to go, will prob back tonight if you want to chat about it | 18:16 |
| *** derekh has quit IRC | 18:16 | |
| vishy | chmouel: can you paste the whole log ? | 18:18 |
| vishy | chmouel: looks like something might have happened earlier | 18:18 |
| *** AlanClark has joined #openstack-dev | 18:21 | |
| *** jeblair has joined #openstack-dev | 18:23 | |
| *** jog0 has quit IRC | 18:25 | |
| *** Ryan_Lane has joined #openstack-dev | 18:28 | |
| *** devananda has quit IRC | 18:28 | |
| *** hhoover has joined #openstack-dev | 18:28 | |
| *** cp16net_ has quit IRC | 18:28 | |
| *** dubsquared has joined #openstack-dev | 18:33 | |
| *** cp16net_ has joined #openstack-dev | 18:34 | |
| *** cp16net_ has left #openstack-dev | 18:34 | |
| *** mnewby has joined #openstack-dev | 18:35 | |
| *** mnewby has quit IRC | 18:35 | |
| *** mnewby has joined #openstack-dev | 18:36 | |
| *** bepernoot has joined #openstack-dev | 18:41 | |
| *** btorch has joined #openstack-dev | 18:42 | |
| dtroyer | notmyname: any chance you can have a look at https://review.openstack.org/5320 regarding the CLI changes we talked about yesterday? | 18:43 |
| notmyname | yes. we'll get to it. there's a lot to do :-) | 18:43 |
| notmyname | dtroyer: you don't have any concerns about releasing one thing in the essex release and then having to always support a migration path? (the underscores in the command names) | 18:44 |
| notmyname | since it's a client-facing change, it needs to be very stable (ie never change) | 18:44 |
| notmyname | I don't have major concerns about this patch, but I do have concerns that it will keep changing. so let's do it right now | 18:45 |
| dtroyer | if it's an userscore vs dash thing, I've lost that argument multiple times…I like dashes myself. | 18:45 |
| dtroyer | this is what the other PTL' | 18:46 |
| btorch | why change those arguments to have os_ seems silly | 18:46 |
| dtroyer | btorch: due to other CLIs that need to present usernames, tenants, passwords, etc as part of the data not as part of the authentication | 18:46 |
| btorch | whoever is already using scripts that use swift tool will break stuff | 18:47 |
| dtroyer | who is using a so-ar undocumented command-line arg? | 18:47 |
| notmyname | btorch: it's not breaking the existing things, just adding to them as an option | 18:48 |
| Glace_ | notmyname: do you consider the new builder.py (ring builder) from 1.4.7 stable? ;) | 18:48 |
| dtroyer | all of the original gas are still present, this is to bring swift in line with all of the other clients regarding keystone authentication | 18:48 |
| dtroyer | s/gas/args/ | 18:48 |
| notmyname | dtroyer: but that's my point about changing the options | 18:48 |
| Glace_ | sorry wrong channel | 18:48 |
| dtroyer | notmyname: I understand, I don't see any reason to change them again. we're only doing this because of some conflicts that arose late | 18:49 |
| dtroyer | the os_ prefix should make them unique enough to prevent that happening again | 18:49 |
| notmyname | Glace_: yes, I consider all of swift that is released to be stable :-) | 18:50 |
| *** deshantm has quit IRC | 18:50 | |
| *** jog0 has joined #openstack-dev | 18:51 | |
| *** belmoreira has quit IRC | 18:54 | |
| notmyname | anotherjesse: should weigh in here (maybe?) | 18:55 |
| anotherjesse | notmyname: reading scrollback | 18:56 |
| *** joesavak has quit IRC | 18:56 | |
| notmyname | anotherjesse: I'm not concerned about the change itself, just that there is a change and may be another, especially after essex | 18:56 |
| *** jakedahn_zz is now known as jakedahn | 18:58 | |
| anotherjesse | notmyname: I like - | 19:00 |
| notmyname | (me too) | 19:00 |
| anotherjesse | notmyname: I like having common-ness between projects | 19:00 |
| anotherjesse | we can either: have vishy+notmyname+bcwaldon+heckj+devcamcar make an executive decision | 19:01 |
| anotherjesse | or email the mailing list | 19:01 |
| heckj | wha? huh? | 19:01 |
| * heckj reads back | 19:01 | |
| anotherjesse | heckj: should it be --os-user-name or --os_user_name | 19:01 |
| *** hhoover has left #openstack-dev | 19:02 | |
| heckj | Oh hell, that again /grumble/ | 19:02 |
| bcwaldon | make it match some established project | 19:02 |
| bcwaldon | I dont care which | 19:02 |
| heckj | I don't really care which, but the topic seems to have become religious | 19:02 |
| anotherjesse | heckj: notmyname brings up that - we might end up having to change | 19:02 |
| bcwaldon | its easier to type with hyphens | 19:02 |
| vishy | chmouel: responded, will try and get it working later | 19:02 |
| anotherjesse | bcwaldon: yeah, I was typing "man foo" for many values of foo to see what multi-word args are usually | 19:02 |
| anotherjesse | dtroyer: who had a strong opinion about using _ | 19:02 |
| notmyname | heckj: the hard part is that we can't change it easily after essex because we'll have to maintain backwards compat | 19:02 |
| dtroyer | they are almost _always_ '-' | 19:02 |
| bcwaldon | then hyphens win | 19:03 |
| *** vincentricci has quit IRC | 19:03 | |
| dtroyer | are we talking about the --os-* only or all long options? | 19:03 |
| bcwaldon | all | 19:03 |
| *** bengrue has joined #openstack-dev | 19:04 | |
| smoser | oh, wait, the bikeshed is clearly best if it is purple | 19:04 |
| bcwaldon | smoser: get out of here | 19:04 |
| anotherjesse | smoser: heh | 19:04 |
| notmyname | smoser: I don't care what color it is, just that we don't repaint it | 19:05 |
| smoser | but i'm almost certain i've probably sworn at every project that uses _ in long argument names. | 19:05 |
| smoser | :) | 19:05 |
| anotherjesse | so we go with _ for now and then in folsom we will end up supporting both | 19:05 |
| bcwaldon | thats such a pain in the ass to support both | 19:05 |
| smoser | i've got a wishlist for clients in folsum | 19:06 |
| notmyname | if we need an executive decision, let's go with "-" | 19:06 |
| dtroyer | I'd rather just write new clients in folsum that have a more consistent command structure too | 19:06 |
| *** byeager has quit IRC | 19:06 | |
| dtroyer | but I don't get a vote.. ;) | 19:06 |
| *** anotherjesse is now known as anotherjesse_zz | 19:07 | |
| smoser | my big wishlist for command line clients is output formatting | 19:09 |
| smoser | i dont knwo which is worse | 19:09 |
| smoser | * parsing an ascii art table | 19:09 |
| smoser | * being asked if you want a client to grab the next page of results (after like 10 results) | 19:09 |
| ironcame12 | dtroyer: i wrote this client and tried to model it after the git command line: https://metacpan.org/module/oscompute | 19:11 |
| ironcame12 | it's definitely not full featured, just have been adding to it incrementally in my free time | 19:12 |
| dtroyer | ironcame12: that's exactly what I did too in https://github.com/dtroyer/drstack using the native client API layers. and a rigorous verb-object command structure | 19:13 |
| ironcame12 | smoser: and its ouput is tab delimted, so you can pipe results into other tools | 19:13 |
| smoser | ironcame12, WHOOO :) | 19:13 |
| smoser | amazon's ec2 api tools aren't great | 19:13 |
| ironcamel | smoser: thanks | 19:13 |
| smoser | but one thing they did right is tab delimit and they never changed the order (ie, they've only ever added fields) | 19:14 |
| ironcamel | dtroyer: great minds think alike :) | 19:14 |
| smoser | (random rant: https://forums.aws.amazon.com/thread.jspa?threadID=81014 ) | 19:14 |
| *** roge has joined #openstack-dev | 19:16 | |
| ironcamel | dtroyer: actually, i went the noun-verb approach, since i was familiar with git | 19:17 |
| ironcamel | like: git remote add blah | 19:17 |
| *** misheska has joined #openstack-dev | 19:17 | |
| dtroyer | ironcamel: I guess my VMS is showing… :( | 19:17 |
| *** misheska has quit IRC | 19:17 | |
| ironcamel | VMS? | 19:17 |
| dtroyer | thanks, I feel even older now | 19:17 |
| ironcamel | hehe, sorry | 19:17 |
| *** misheska has joined #openstack-dev | 19:18 | |
| *** dolphm has joined #openstack-dev | 19:18 | |
| *** ayoung has joined #openstack-dev | 19:19 | |
| *** msinhore has joined #openstack-dev | 19:20 | |
| *** anotherjesse_zz is now known as anotherjesse | 19:20 | |
| msinhore | hello all | 19:21 |
| anotherjesse | dtroyer: so notmyname's decision to use - instead of _ is landing? | 19:22 |
| msinhore | I just know upload a openstack-nova.spec file for centos in github with support to xenapi package to be installed in xenserver and xcp host: https://github.com/timeredbull/openstack-essex-4-rpms/blob/master/openstack-nova.spec | 19:22 |
| msinhore | s/know/now/ | 19:22 |
| *** roge has quit IRC | 19:22 | |
| notmyname | anotherjesse: well, it will be up to someone else to submit the patches ;-) | 19:23 |
| dtroyer | anotherjesse: that is a huge change for all of the other clients…especially to support both styles for essex... | 19:23 |
| *** roge has joined #openstack-dev | 19:23 | |
| msinhore | if someone are interested in help testing or suggesting code, I will be glad | 19:23 |
| anotherjesse | dtroyer: so your vote is _ due to minimizing risk - since we will want to redo them later | 19:23 |
| *** adjohn has quit IRC | 19:23 | |
| *** markmcclain has joined #openstack-dev | 19:23 | |
| dtroyer | anotherjesse: yes. I do personally prefer - though | 19:24 |
| heckj | anotherjesse, dtroyer, notmyname, etc…: I'd prefer to change as little as possible. We're still nailing down the docs, and every time we tweak these minimal things, there's the change to the code and then the change to the docs, which is often more intrusive. | 19:24 |
| dtroyer | and that's why I want to have a do-over in folsum for all of the CLI interfaces | 19:24 |
| anotherjesse | notmyname: are you ok with that .. you will be allowed to be :( about it | 19:24 |
| notmyname | heh | 19:24 |
| heckj | I'm totally good with a do-over in folsom to switch it up | 19:24 |
| anotherjesse | having the project names in the clients is confusing anyway … we tell people it is compute but then they have to know to use the cli tool "nova" | 19:25 |
| notmyname | if things are going to change, why worry about this patch landing now? | 19:26 |
| anotherjesse | notmyname: because we have users now | 19:26 |
| *** roge has quit IRC | 19:26 | |
| notmyname | so what are they doing today? | 19:26 |
| anotherjesse | honestly - there hasn't been a unified auth story until essex | 19:26 |
| Vek | I assume the "too many links" problem with glance approvals is known about? | 19:26 |
| anotherjesse | diablo keystone was delayed | 19:27 |
| anotherjesse | so even the half-integrated version wasn't used in many diablo installs | 19:27 |
| *** roge has joined #openstack-dev | 19:27 | |
| ironcame12 | can someone please point me to the proper zodiac repo. these repo's don't seem to be active: https://github.rackspace.com/daryl-walleck/Zodiac https://github.com/dwalleck/zodiac | 19:33 |
| ironcame12 | sorry, wrong window | 19:33 |
| ironcame12 | but will still accept an answer if anyone knows :) | 19:34 |
| anotherjesse | ironcame12: it became github.com/openstack/tempest | 19:34 |
| ironcame12 | anotherjesse: thanks! | 19:34 |
| jaypipes | jeblair: any chance you could look into https://jenkins.openstack.org/job/gate-glance-python27/100/console and give me some ideas as to why it's failing? | 19:35 |
| jeblair | jaypipes: looking | 19:35 |
| ironcame12 | so anotherjesse said it is github.com/openstack/tempest | 19:35 |
| ironcame12 | crap! | 19:35 |
| ironcame12 | i suck at irssi | 19:35 |
| anotherjesse | anotherjesse: np | 19:36 |
| *** rohita has joined #openstack-dev | 19:36 | |
| rmk | trystack event log is handy -- where did that and the billing stuff come from? | 19:37 |
| *** hitesh__ has quit IRC | 19:37 | |
| anotherjesse | rmk: both are in github - looking for repos | 19:37 |
| rmk | ah nice | 19:37 |
| rmk | which repos are they in now? :) | 19:38 |
| anotherjesse | rmk: https://github.com/ntt-pf-lab - billing and utils | 19:38 |
| rmk | awesome thanks | 19:38 |
| vishy | vek, comstud: updated https://review.openstack.org/#patch,sidebyside,5342 | 19:38 |
| vishy | er: https://review.openstack.org/#change,5342 | 19:39 |
| jeblair | jaypipes: cause is test files leaking in /tmp. cleaning up. we have an open bug to configure tmpwatch, but no one has grabbed that yet. | 19:40 |
| jaypipes | jeblair: gotcha. thx | 19:46 |
| *** heckj has quit IRC | 20:02 | |
| *** anotherjesse is now known as anotherjesse_zz | 20:05 | |
| *** eglynn_ has quit IRC | 20:06 | |
| *** roge has quit IRC | 20:06 | |
| *** eglynn_ has joined #openstack-dev | 20:06 | |
| *** stuntmachine has quit IRC | 20:12 | |
| *** anotherjesse_zz is now known as anotherjesse | 20:16 | |
| *** rkukura has quit IRC | 20:23 | |
| *** bepernoot has quit IRC | 20:26 | |
| *** zigo has quit IRC | 20:29 | |
| jog0 | as we are getting ready for RC1, what is the proper capitalization of 'openstack' OpenStack? The code base has all kinds of capitalization: Openstack, openstack and OpenStack | 20:30 |
| *** bepernoot has joined #openstack-dev | 20:30 | |
| anotherjesse | smoser: would you like to look at this non-purple patch? https://review.openstack.org/#change,5363,patchset=2 it reverts the logic changes I made to metadata but keeps the caching layer | 20:30 |
| smoser | did you *want* me to reject it? i clearly stated purple was the best. | 20:31 |
| smoser | i'll give it a spin | 20:31 |
| anotherjesse | the first call takes 400ms, but calls afterwards take < 5ms | 20:31 |
| *** bepernoot has quit IRC | 20:31 | |
| anotherjesse | I had to move forward bcwaldon's patch for image_id | 20:32 |
| *** roge has joined #openstack-dev | 20:38 | |
| *** bencherian has quit IRC | 20:39 | |
| anotherjesse | smoser: what is a good time for a cloud-init run? | 20:40 |
| *** bepernoot has joined #openstack-dev | 20:40 | |
| anotherjesse | cloud-init start running: [snip] up 3.85 seconds / cloud-init boot finished [snip] Up 5.33 seconds ; 1.5s total | 20:41 |
| smoser | ah. | 20:41 |
| smoser | a more direct test of the MD speed would be just with boto.utils.get_instance_metadata() | 20:41 |
| smoser | but 1.5 seconds is not bad. | 20:42 |
| smoser | the biggest issue i think that people had was that it got worse with size. | 20:42 |
| smoser | siez of cluster | 20:42 |
| anotherjesse | smoser: do you guys run an metadata api server per compute node? | 20:43 |
| smoser | no. canonistack did not. | 20:44 |
| smoser | but it was 5 machines. | 20:44 |
| anotherjesse | smoser: next step is going to be launching hundreds of VMs and reboot them all at the same time… it *shouldn't* be that bad | 20:45 |
| smoser | right. | 20:45 |
| smoser | or just executing a crawl on all of them | 20:45 |
| smoser | (ie, no need to reboot) | 20:45 |
| anotherjesse | smoser: for larger deploys it now supports putting it in memcache - if you have multiple metadata api servers it would then share the speed hit on the first api call | 20:45 |
| smoser | so, cirros's ec2metadata looked improved with your patch | 20:47 |
| smoser | (the output) | 20:47 |
| smoser | ie, http://bazaar.launchpad.net/~smoser/cirros/trunk/view/head:/src/usr/bin/ec2metadata | 20:47 |
| smoser | just running that in an instance | 20:48 |
| anotherjesse | smoser: executing python -c 'import boto.utils, pprint; pprint.pprint(boto.utils.get_instance_metadata())' takes 0.139s | 20:48 |
| smoser | so, on canonical's openstack right now, that takes 2.3 seconds. | 20:48 |
| smoser | 2.24 | 20:48 |
| anotherjesse | each metadata call is doing a RPC call over the queue to the network service to map IP to instance | 20:49 |
| anotherjesse | takes at least 100ms to do that | 20:49 |
| smoser | right. the rpc call is what was worst. | 20:51 |
| smoser | so... | 20:51 |
| smoser | this sniffs ok to me. | 20:51 |
| smoser | instances boot, and no complaints that i see. | 20:52 |
| *** eglynn__ has joined #openstack-dev | 20:53 | |
| smoser | anotherjesse, so can you confirm... | 20:54 |
| smoser | basically the first call is going to suck performance wise | 20:54 |
| smoser | as its still going to fill up all the data and everything | 20:54 |
| *** dprince has joined #openstack-dev | 20:55 | |
| anotherjesse | smoser: yes | 20:56 |
| smoser | oh wait, do subsequent calls hit the RPC call ? | 20:56 |
| smoser | i think yes. | 20:56 |
| anotherjesse | hmm? | 20:56 |
| *** eglynn_ has quit IRC | 20:56 | |
| anotherjesse | get_metadta call checks the cache first thing | 20:56 |
| anotherjesse | at the end of the function it caches for 15 seconds | 20:56 |
| *** vincentricci has joined #openstack-dev | 20:56 | |
| *** adjohn has joined #openstack-dev | 20:57 | |
| anotherjesse | smoser: here is my nova-api log http://pastie.org/pastes/3596391/text | 20:58 |
| anotherjesse | for a single vm calling the get_instance_metadata twice | 20:59 |
| *** Tushar has joined #openstack-dev | 21:00 | |
| davidkranz | jaypipes: What do lines that just say "ERROR" in the middle of tempest output mean? | 21:00 |
| *** bepernoot has quit IRC | 21:01 | |
| jaypipes | davidkranz: lol, not sure. | 21:01 |
| smoser | anotherjesse, well, hopefully at some point we can do better. | 21:01 |
| smoser | but for now, the big issue was cloud-init's stupid 2 second time out | 21:01 |
| smoser | and we were hitting that, it was taking 5 seconds for a single response. | 21:01 |
| jaypipes | davidkranz: I would suspect a stray print has made its way into the code | 21:02 |
| davidkranz | jaypipes: OK, I'll check. | 21:04 |
| *** martine has quit IRC | 21:04 | |
| *** RobertLaptop has joined #openstack-dev | 21:04 | |
| *** RobertLaptop has left #openstack-dev | 21:05 | |
| *** zzed has joined #openstack-dev | 21:05 | |
| *** eglynn__ has quit IRC | 21:05 | |
| *** eglynn__ has joined #openstack-dev | 21:05 | |
| *** markvoelker has joined #openstack-dev | 21:06 | |
| *** RobertLaptop has joined #openstack-dev | 21:07 | |
| *** RobertLaptop has left #openstack-dev | 21:07 | |
| rmk | anotherjesse: if you have a moment -- been working on adding this dash_billing component. I've got it installed in my venv, added it to the INSTALLED_APPS array in settings.py, but nothing shows up. Is there another step I am missing? | 21:08 |
| anotherjesse | rmk: in diablo? | 21:08 |
| rmk | yep | 21:08 |
| rmk | I can runserver with no errors loading the module but I don't see anything new in the dash | 21:09 |
| anotherjesse | rmk: join #trystack | 21:09 |
| *** tryggvil__ has quit IRC | 21:10 | |
| *** eglynn__ has quit IRC | 21:10 | |
| bcwaldon | jeblair: any idea why I can't pass the python 2.6/7 tests in jenkins for glance? | 21:10 |
| *** eglynn__ has joined #openstack-dev | 21:10 | |
| jeblair | bcwaldon: link to job? | 21:12 |
| jeblair | or change? | 21:12 |
| *** donaldngo_hp has joined #openstack-dev | 21:12 | |
| bcwaldon | jeblair: all glance reviews have failures | 21:13 |
| bcwaldon | jeblair: https://review.openstack.org/#q,status:open+project:openstack/glance,n,z | 21:13 |
| jeblair | https://jenkins.openstack.org/job/gate-glance-python26/98/console | 21:14 |
| jeblair | is failing tests | 21:14 |
| jeblair | https://jenkins.openstack.org/job/gate-glance-python27/100/console is because glance tests leak files in /tmp, and the directory filled up on the slave, i've since corrected that | 21:15 |
| jeblair | same for https://jenkins.openstack.org/job/gate-glance-python27/103/console | 21:15 |
| jeblair | and for https://jenkins.openstack.org/job/gate-glance-python27/102/console | 21:16 |
| bcwaldon | jeblair: ok, so can I just reapprove the last three? | 21:17 |
| jeblair | yes, those should test correctly then | 21:17 |
| bcwaldon | ok, thanks a bunch | 21:17 |
| jeblair | as for 5346 | 21:17 |
| jeblair | you should be aware that at jaypipes request we configured S3 and swift credentials for some functional tests | 21:17 |
| *** danwent has quit IRC | 21:18 | |
| donaldngo_hp | jeblair: I cherry-picked something from a branch proposed against trunk into a local repo branched from stable/diablo. everything looks cool up until gerrit rejecting the commit. The output is here: http://paste.openstack.org/show/9432/ | 21:18 |
| *** dolphm has quit IRC | 21:18 | |
| jeblair | which means the unit tests could be subject to externally caused failures on those tests; just keep that in mind. | 21:18 |
| bcwaldon | jeblair: yes, I think we've been running into those failures | 21:19 |
| jeblair | bcwaldon: i'm not sure i'm a fan of depending on external test resources we don't control for gating. :( at least with the devstack gates we can improve our own code and yell^W engage the assistance of our cloud providers. i guess let's see how this works out. we can probably stand up a swift for functional testing. an S3 might be harder. :) | 21:21 |
| bcwaldon | yeah, I hear ya | 21:22 |
| jeblair | donaldngo_hp: so that error means... | 21:22 |
| jeblair | donaldngo_hp: that your change had "Change-Id: I025c5728f0a0a35d9b42cb70e1da9b822f4e6553" in the commit message | 21:22 |
| jeblair | which has been abandoned, so gerrit won't accept an update to it. | 21:22 |
| jeblair | (if you upload a change with a Change-Id, and gerrit already has a change with that id, it assumes you're uploading a new patchset for the existing change) | 21:23 |
| jaypipes | jeblair: the swift tests in glance test against cloud files. so you can yell at the cloud provider all you want... for the S3 tests, I suppose we could stand up Swift + Swift3, but I'd rather use the canonical source in AWS S3 | 21:24 |
| jeblair | you probably just want to 'git commit -a --amend' the change, delete the Change-Id line, and 'git review' again. | 21:24 |
| *** dolphm has joined #openstack-dev | 21:24 | |
| *** dolphm has quit IRC | 21:24 | |
| *** markvoelker has quit IRC | 21:25 | |
| jeblair | jaypipes: i said "engage" not yell. :) are these S3 failures? https://jenkins.openstack.org/job/gate-glance-python26/98/console | 21:25 |
| bcwaldon | jeblair: consistently seing this failure -> https://jenkins.openstack.org/job/gate-glance-python26/99/console | 21:28 |
| jeblair | but build 100 worked | 21:29 |
| jaypipes | jeblair: I don't think those are S3 errors, no... | 21:29 |
| jeblair | ok, so i think we're running the tests correctly; i'm counting on you guys being able to diagnose the error from the test output. | 21:32 |
| jog0 | bcwaldon: what is nova-cert used for? | 21:32 |
| donaldngo_hp | jeblair: the change id keeps on coming back after i do a git commit -a --amend and delete the change id | 21:32 |
| jeblair | donaldngo_hp: it gets replaced with a new change id, that's what you want. | 21:32 |
| *** markmcclain has quit IRC | 21:33 | |
| donaldngo_hp | jebliar: still getting error: failed to push some refs... | 21:33 |
| donaldngo_hp | jeblair: btw im doing a git review gerrit/stable/diablo | 21:34 |
| donaldngo_hp | the git review worked and i see my review proposed for master | 21:35 |
| jeblair | donaldngo_hp: is something still failing? | 21:35 |
| *** statik has quit IRC | 21:37 | |
| donaldngo_hp | jeblair: yea its the same error: http://paste.openstack.org/show/9439/ | 21:38 |
| *** andrewsmedina has quit IRC | 21:38 | |
| jeblair | donaldngo_hp: does "I025c5728f0a0a35d9b42cb70e1da9b822f4e6553" show up in 'git log'? | 21:41 |
| donaldngo_hp | jeblair: yea i see it as a change ID for one of the commits by jay | 21:43 |
| jaypipes | jeblair: there is a bug in the copy_from_ test case. for now, would you mind disabling those environment variables so those tests do not run? | 21:44 |
| *** lts has quit IRC | 21:45 | |
| jeblair | jaypipes: done | 21:45 |
| *** andrewsmedina has joined #openstack-dev | 21:46 | |
| *** eglynn__ has quit IRC | 21:47 | |
| *** eglynn has joined #openstack-dev | 21:47 | |
| jaypipes | jeblair: tyvm | 21:48 |
| jaypipes | eglynn: ah, perfect timing :) | 21:48 |
| *** markmcclain has joined #openstack-dev | 21:48 | |
| jaypipes | eglynn: mind taking a look at https://bugs.launchpad.net/glance/+bug/955527 ? | 21:48 |
| uvirtbot | Launchpad bug 955527 in glance "copy_from test case logic is invalid" [High,Confirmed] | 21:48 |
| jaypipes | bcwaldon: I think you should be able to clear the backlog at this point for glance reviews. | 21:48 |
| * eglynn looking ... | 21:48 | |
| jaypipes | eglynn: cheers mate | 21:48 |
| bcwaldon | jaypipes: kk | 21:48 |
| eglynn | jaypipes: a-ha, because of the deferred delete logic in the glance API layer, right? | 21:51 |
| eglynn | jaypipes: would be equivalent to just set the scrub_time config to zero? | 21:52 |
| eglynn | jaypipes: nope, I'm talking rubbish, nowt to do with deferred delete .... | 21:54 |
| anotherjesse | adam_g: I think there are bugs to associate your patch with | 21:55 |
| anotherjesse | adam_g: I think https://bugs.launchpad.net/nova/+bug/955351 -- so you would add in the message part of your comment "bug #955351" to auto-link to launchpad / cause all the magic to occur | 21:56 |
| uvirtbot | Launchpad bug 955351 in nova "Permission denied when getting console output" [High,Triaged] https://launchpad.net/bugs/955351 | 21:56 |
| uvirtbot | Launchpad bug 955351 in nova "Permission denied when getting console output" [High,Triaged] | 21:56 |
| *** Glace_ has quit IRC | 21:56 | |
| *** statik has joined #openstack-dev | 21:58 | |
| *** statik has joined #openstack-dev | 21:58 | |
| *** dolphm has joined #openstack-dev | 22:01 | |
| *** statik has quit IRC | 22:02 | |
| *** eglynn has quit IRC | 22:03 | |
| *** eglynn_ has joined #openstack-dev | 22:03 | |
| *** eglynn__ has joined #openstack-dev | 22:04 | |
| adam_g | anotherjesse: ah, thanks. was going to open one myself in a min and amend, shoulda checked first. thanks | 22:06 |
| *** eglynn has joined #openstack-dev | 22:06 | |
| eglynn | jaypipes: but the fact that the sunsequent get of the copied image ID fails with a 404 indicates either (a) it isn't really independent of the existence of the original or (b) swift is intermittently broken | 22:06 |
| eglynn | jaypipes: so the point of that delete in the test is to ensure that the copied image (in swift) is independent of the original image (in s3) | 22:07 |
| *** eglynn_ has quit IRC | 22:07 | |
| eglynn | jaypipes: I think it should be OK for the deletion of the *original* image to be mediated by the API layer, to avoid the test being directly boto-aware | 22:08 |
| *** eglynn__ has quit IRC | 22:09 | |
| *** zzed has quit IRC | 22:11 | |
| *** shevek_ has joined #openstack-dev | 22:11 | |
| *** statik has joined #openstack-dev | 22:11 | |
| *** statik has joined #openstack-dev | 22:11 | |
| anotherjesse | nova core - the following bugs are blocking using/debugging nova with UEC images: https://review.openstack.org/#change,5363 (ec2 metadata is broken) and https://review.openstack.org/#change,5371 (unable to get log) | 22:12 |
| *** eglynn_ has joined #openstack-dev | 22:12 | |
| *** eglynn has quit IRC | 22:13 | |
| *** eglynn_ has quit IRC | 22:14 | |
| *** eglynn has joined #openstack-dev | 22:15 | |
| eglynn | jaypipes: just noticed that the mysterious 404 returned by swift afflicts other tests intermittently | 22:15 |
| eglynn | jaypipes: e.g TestS3.test_remote_image in https://jenkins.openstack.org/job/gate-glance-python26/85/console | 22:15 |
| * eglynn begins to suspect gremlins in swift-land ... | 22:15 | |
| *** sannes has quit IRC | 22:17 | |
| anotherjesse | russellb: does pam still work? | 22:18 |
| eglynn | jaypipes: gotta drop off for an hour or so, back later ... | 22:24 |
| russellb | anotherjesse: no idea ... i tested it manually/directly, heh | 22:27 |
| *** msinhore has quit IRC | 22:28 | |
| *** dprince has quit IRC | 22:30 | |
| anotherjesse | russellb: we started with PAM way back in the day, and no one seemed interested … we should try it out and if it doesn't work remove or fix | 22:31 |
| russellb | sounds good to me | 22:31 |
| *** andrewsmedina has quit IRC | 22:31 | |
| russellb | i was mostly interested from a "if it works elsewhere, i want it to work on Fedora" perspective | 22:31 |
| russellb | removing it would solve that criteria too if it doesn't work and nobody cares .. | 22:32 |
| russellb | but i can give it a spin and see | 22:32 |
| anotherjesse | wonder if ayoung thinks about pam | 22:32 |
| anotherjesse | s/if/what/ | 22:32 |
| *** bsza has quit IRC | 22:34 | |
| vishy | comstud: can you rekick https://review.openstack.org/#change,5342 | 22:37 |
| vishy | nm someone got it | 22:37 |
| *** littleidea has quit IRC | 22:38 | |
| *** johnpostlethwa-1 is now known as johnpostlethwait | 22:40 | |
| *** AlanClark has quit IRC | 22:46 | |
| comstud | yea | 22:48 |
| *** rkukura has joined #openstack-dev | 22:52 | |
| anotherjesse | jeblair: no nodes error :-/ | 22:52 |
| jeblair | grr | 22:52 |
| anotherjesse | https://jenkins.openstack.org/job/gate-integration-tests-devstack-vm/2566/ | 22:53 |
| *** Tushar has quit IRC | 22:53 | |
| jeblair | pvo: https://jenkins.openstack.org/user/corvus/my-views/view/Integration/job/devstack-launch-vms/11527/console | 22:53 |
| jeblair | pvo: 400 Bad Request 422 Unprocessable Entity: There was an error adding the requested server.; | 22:54 |
| jeblair | pvo: any chance you can help with that? | 22:54 |
| *** dtroyer is now known as dtroyer_zzz | 23:00 | |
| jeblair | (filed a ticket, talking to rs cloud support) | 23:01 |
| adam_g | vishy: you reckon https://review.openstack.org/#change,5342 is a fix for https://bugs.launchpad.net/ubuntu/+source/nova/+bug/954692 as well? | 23:08 |
| uvirtbot | Launchpad bug 954692 in nova "cannot detach volume from terminated instance" [Medium,New] | 23:08 |
| vishy | adam_g: no | 23:11 |
| vishy | volumes should be detached automatically when you terminate | 23:13 |
| vishy | if this isn't happening it is a separate bug | 23:13 |
| *** deshantm has joined #openstack-dev | 23:14 | |
| *** dtroyer_zzz is now known as dtroyer | 23:15 | |
| *** dolphm has quit IRC | 23:15 | |
| *** paulormg has quit IRC | 23:18 | |
| *** kbringard has quit IRC | 23:20 | |
| jeblair | anotherjesse, pvo: rs support says: "We are working to get this resolved for you, our technicians are engaged to get this back up and running as soon as possible" | 23:21 |
| jeblair | (ticket #349412 in case any rackers are interested) | 23:21 |
| pvo | jeblair: whats going on? | 23:21 |
| pvo | k, I'll look | 23:21 |
| *** rods has quit IRC | 23:23 | |
| *** dolphm has joined #openstack-dev | 23:24 | |
| pvo | jeblair: engineering is investigating that issue now. | 23:25 |
| jeblair | groovy | 23:26 |
| anotherjesse | jeblair: could we catch that exception and put "FAILED (to get cloud server)" rather than FAILED | 23:34 |
| *** msinhore has joined #openstack-dev | 23:35 | |
| *** msinhore has quit IRC | 23:35 | |
| *** dolphm has quit IRC | 23:41 | |
| *** dolphm_ has joined #openstack-dev | 23:43 | |
| *** littleidea has joined #openstack-dev | 23:43 | |
| *** ipl31 has quit IRC | 23:45 | |
| *** benner has quit IRC | 23:45 | |
| *** hazmat has quit IRC | 23:45 | |
| *** termie has quit IRC | 23:45 | |
| *** ahale has quit IRC | 23:45 | |
| *** benner has joined #openstack-dev | 23:45 | |
| *** ahale has joined #openstack-dev | 23:46 | |
| *** termie has joined #openstack-dev | 23:46 | |
| *** ipl31 has joined #openstack-dev | 23:46 | |
| *** hazmat has joined #openstack-dev | 23:46 | |
| *** hazmat has quit IRC | 23:46 | |
| *** hazmat has joined #openstack-dev | 23:46 | |
| *** mszilagyi has quit IRC | 23:47 | |
| *** donaldngo_hp has quit IRC | 23:48 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!