| aNupoisc | sdake: Mech422: is this the correct repo for docker "deb https://apt.dockerproject.org/repo ubuntu-trusty main" | 00:00 |
|---|---|---|
| jrich523 | sdake: so im looking at the multinode inventory and it has section (much like the all-in-one) but has, fake? names in most of them | 00:00 |
| aNupoisc | is this the correct repo for docker "deb https://apt.dockerproject.org/repo ubuntu-trusty main" | 00:00 |
| jrich523 | do i just blow them away? or will that uninstall/remove stuff? | 00:01 |
| sdake | jrich523 yes blow away the fake names | 00:01 |
| jrich523 | k | 00:01 |
| sdake | they are just samples | 00:01 |
| sean-k-mooney | sbezverk: just checked the ovs agent code and your write for the pyical bridge is dose not create them. | 00:01 |
| Mech422 | sean-k-mooney: sbezverk: I have to manually create br-ex as well | 00:02 |
| sean-k-mooney | sbezverk: i can change it in the agent if you like. its litrally moving 2 lines | 00:03 |
| sbezverk | sean-k-mooney: I just learned today about init-container in kube I will use it with neutron-ovs-agent to create br-ex before the agent process starts | 00:03 |
| Mech422 | sean-k-mooney: sbezverk when kolla starts up ovs, it starts with an empty DB - so I stop kolla from starting the ovs-db stuff | 00:03 |
| sbezverk | sean-k-mooney: with init containers you do need kolla bootstraping | 00:03 |
| sean-k-mooney | sbezverk: actully the correct behavior is to create all the bridge before ovs-vswitchd starts | 00:03 |
| sean-k-mooney | Mech422: that was intentional | 00:04 |
| Mech422 | sean-k-mooney: it blows up ALL my host networking, since my hosts are ovs based :-P | 00:04 |
| Mech422 | insta-brick! | 00:04 |
| sbezverk | sean-k-mooney: oh, interesting in this case I will use init with vswitchd | 00:04 |
| sean-k-mooney | Mech422: a well you are not ovs deployed with kolla | 00:05 |
| sbezverk | sean-k-mooney: cool, good to know, you saved me day of troubleshooting ;-) | 00:05 |
| openstackgerrit | Wei Cao proposed openstack/kolla: Add solum container https://review.openstack.org/355408 | 00:05 |
| Mech422 | sean-k-mooney: no - I'm not , the bare metal uses OVS based networking | 00:05 |
| jrich523 | so i got this error when i run the precheck on that config file, however... it seems to be ok? or maybe its suppose to fail? http://paste.openstack.org/show/561553/ | 00:06 |
| Mech422 | sean-k-mooney: so I sorta need to install it before I install kolla :-) | 00:06 |
| jrich523 | it might be trying to fail, since it is able to ping, i ran my internal/external networks on the same subnet... because.. well... im creating it all on my pc with vmware workstation | 00:06 |
| jrich523 | i also manually tested the pings, which are fine | 00:07 |
| sean-k-mooney | sbezverk: well for kernel ovs its not as required but for ovs-dpdk you cannont add interfaces once the ovs-vswitchd has started so all bridge and interface config is done after the ovsdb is started but before the ovs-vswitchd is started | 00:07 |
| sean-k-mooney | Mech422: so out of interesst what do you use ovs on the host for? | 00:09 |
| sean-k-mooney | Mech422: are you using it for your management and api networks | 00:09 |
| sdake | jrich523 what is .31 | 00:09 |
| jrich523 | vip | 00:09 |
| Mech422 | sean-k-mooney: to replicate our production setup... yeah, we have hostmgmt, and all openstack networks in ovs | 00:10 |
| jrich523 | 30 is the controller host, 31 is vip, 21 is this new host used for compute | 00:10 |
| Mech422 | sean-k-mooney: our original setup uses liberty, and ovs with bonded 10G nics | 00:10 |
| sdake | what node are you deploing rom? | 00:10 |
| jrich523 | its a CentOS7 box | 00:10 |
| sdake | which ip? | 00:10 |
| jrich523 | which is the same as the controller | 00:11 |
| jrich523 | 21 | 00:11 |
| sdake | so the controller is where you have this multinode file? | 00:11 |
| jrich523 | yes | 00:11 |
| sean-k-mooney | Mech422: ah makes sense. there is some refactoring around how we deploy ovs that i wanted to do next cycle. it would be interesting if i could make your usecase work | 00:11 |
| sdake | and it contains both .21 and .31 in the invetory - one for controller and one for compute? | 00:11 |
| jrich523 | well, it just contains 21, should i add 31 to all other sections? | 00:11 |
| sdake | if you want to deploy 31, it needs to be in the inventory file | 00:12 |
| sean-k-mooney | Mech422: basiclaly i wanted to split deploying the network backend out of the neutron role. | 00:12 |
| Mech422 | sean-k-mooney: bonus points if our vxlan private networks work :-) | 00:12 |
| jrich523 | its already deployed | 00:12 |
| sdake | if you want to deploy 21 it needs to be in the invetory file | 00:12 |
| jrich523 | i (guess) the all-in-one was done originally (Basically quick start guide) | 00:12 |
| sdake | if you want to deploy 21 and 31 they both need to be in the invetory file | 00:12 |
| sdake | jrich523 ya - run kolla-ansible -i multinode destroy | 00:12 |
| sdake | that will clean up things | 00:12 |
| Mech422 | sean-k-mooney: I should be able to test stuff in the next week or two if you like | 00:13 |
| sdake | oh your using a stable branch | 00:13 |
| sdake | that wont work | 00:13 |
| Mech422 | sean-k-mooney: they got me on cloudstack atm | 00:13 |
| jrich523 | scary | 00:13 |
| * sdake groans | 00:13 | |
| jrich523 | so 31 is deployed correctly and working | 00:13 |
| v1k0d3n | sdake: hey man | 00:13 |
| sean-k-mooney | ew competing could tech :) | 00:13 |
| jrich523 | it has horizon on it and looks good | 00:13 |
| sdake | sup v1k0d3n | 00:13 |
| v1k0d3n | i missed your message earlier. | 00:13 |
| jrich523 | i just want to add a compute node | 00:13 |
| sdake | v1k0d3n i was just complaining about tooth pain and the need for hyddrocodone | 00:13 |
| v1k0d3n | yo | 00:13 |
| v1k0d3n | sorry man. hahaha | 00:13 |
| Mech422 | sean-k-mooney: yeah - oh well..knowledge is money^H^H^H^H^H power | 00:14 |
| sdake | jrich to do that you hvae to run kolla-ansible deplooy | 00:14 |
| sdake | with the correct inventory file | 00:14 |
| v1k0d3n | dude, careful with that stuff. :) | 00:14 |
| Mech422 | sean-k-mooney: though the cloudstack networking is driving me batshit | 00:14 |
| v1k0d3n | sorry your having pain | 00:14 |
| sdake | it was just a joke | 00:14 |
| sean-k-mooney | Mech422: i wont get to look at it until october | 00:14 |
| jrich523 | so if all i want to do is deploy the compute part, do i need the multinode to have the controller info as well since thats already good to go? | 00:14 |
| Mech422 | sean-k-mooney: ahh - I should be able to help/test then :-) | 00:14 |
| jrich523 | use scotch instead, its safer ;) | 00:15 |
| sdake | jrich yes | 00:15 |
| sean-k-mooney | Mech422: but i might ask you what you would need either kolla host or the openvsiwthc role to support for your usecase | 00:15 |
| v1k0d3n | yeah, hey sdake saw the _member_ role is still out there? | 00:15 |
| sdake | jrich523 yes - yo uhave to have everything in the invetory file | 00:15 |
| jrich523 | ok, i'll add the controller node to all the sections and see what blows up | 00:15 |
| jrich523 | speaking of which, i'll snapshot this first lol | 00:15 |
| sdake | jrich523 before doing anything | 00:16 |
| Mech422 | sean-k-mooney: we do ovs bonds on the metal, break out interfaces by vlan for openstack, and use 1 vlan for vxlan private networking...thats about it so far - no lbaas or fwaas | 00:16 |
| sdake | paste the inveotry file | 00:16 |
| sean-k-mooney | Mech422: but ya after figuing out how neutron works i dont think i could handle another could networking stack | 00:16 |
| sdake | i mean before doing a deploy | 00:16 |
| jrich523 | k | 00:16 |
| sdake | jrich523 the reason everything needs to be in the invetry file is so the compute and controller nodes know how to speak to each other | 00:16 |
| sdake | and understand all the databae passwords | 00:16 |
| sdake | and whatnot | 00:16 |
| sean-k-mooney | Mech422: ah that makes sense | 00:16 |
| jrich523 | that makes sense, i guess mostly i thought i'd find some file (be it all-in-one or multinode) that already had that info, because... well.... its already been deployed | 00:17 |
| sdake | sean-k-mooney lol | 00:17 |
| jrich523 | so i was mildly confused that i didnt see that config in any of the inventories | 00:17 |
| sean-k-mooney | Mech422: people always recommended dedicating a vlan to carry all vxlan traffic just havent met anyone who did it till now | 00:17 |
| Mech422 | sean-k-mooney: we have a bunch of 4 node 'utility' boxes deployed at about 80 pops around the world, which they want to 'centrally manage' - its not really a good fit for openstack, but cloudstack has a nice heirachial setup, and eucastack could be a possiblity | 00:17 |
| sdake | jrich523 you may have to clea nup - not sure | 00:17 |
| sdake | jrich523 we will find out together | 00:18 |
| jrich523 | where is the 'running' inventory file? | 00:18 |
| sdake | all-in-one is run by default | 00:18 |
| jrich523 | k | 00:18 |
| jrich523 | oh right, it has localhost all over the place | 00:18 |
| Mech422 | sean-k-mooney: LOL - I went WAY overboard on the vlans...seperate ceph replication network, seperate db repl. network, seperate api network | 00:18 |
| sdake | yup | 00:18 |
| sdake | it may or may not work | 00:18 |
| Mech422 | sean-k-mooney: neteng hated me :-P | 00:18 |
| sdake | as a result | 00:18 |
| jrich523 | which i think is why you told me to use that one the other day | 00:18 |
| sdake | we may need to clean up when going from aio to multinode | 00:19 |
| jrich523 | but i guess it shouldnt matter which i use, provided the localhost (30/31) is all over the place and 21 is in the compute section | 00:19 |
| sdake | but lets try it | 00:19 |
| jrich523 | k | 00:19 |
| sean-k-mooney | Mech422: ah yes i have heard of simlar topologies in the telcoms space for vCPE deployments were tecol want to virtuilse there on prem equipment using openstakc to manage it centrally | 00:19 |
| sdake | paste inventory file | 00:19 |
| sdake | beffore running a deploy | 00:19 |
| jrich523 | yup grabbing it now | 00:19 |
| jrich523 | so i should use .30 right? | 00:20 |
| *** bmace has quit IRC | 00:20 | |
| sean-k-mooney | Mech422: well in the plus side if you dont have expensive switches that can do qos on vxlan traffic you can just apply it to the vxlan vlan | 00:20 |
| jrich523 | for [control] | 00:20 |
| *** bmace has joined #openstack-kolla | 00:20 | |
| sdake | jrich523 I thought 21 was your control node? | 00:21 |
| sean-k-mooney | Mech422: similarly for you ceph network though i dont think i personally would have create a deicated vlan for the dbs | 00:21 |
| jrich523 | no it will be a compute node | 00:21 |
| aNupoisc | Hi guys, when I am trying to install docker-engine it is throwing me error http://pastebin.com/pYKmiYGS | 00:21 |
| jrich523 | 30 is currently an "all-in-one" setup | 00:21 |
| sdake | .31 is your control node? | 00:21 |
| jrich523 | 31 is the vip | 00:21 |
| sdake | yup put 30 in there | 00:22 |
| sdake | the full ip | 00:22 |
| jrich523 | k | 00:22 |
| jrich523 | http://paste.openstack.org/show/561554/ | 00:22 |
| sdake | uder network put 30 | 00:22 |
| jrich523 | removed some comments fore readability | 00:22 |
| sdake | under storage put 30 | 00:22 |
| jrich523 | yeah all over the place | 00:22 |
| jrich523 | and used 21 in Storage and Compute | 00:22 |
| sdake | under compute put 21 | 00:22 |
| sdake | 21 is fine for compute | 00:22 |
| Mech422 | sean-k-mooney: yeah - I got a bit carried away :-) We a CDN network, so sorta similiar to telco - lots of small caches of equipment all over, with a few 'major' caches centrally located | 00:22 |
| sdake | are you using storage such ass cinder? | 00:22 |
| aNupoisc | is the distribution I am using correct? | 00:23 |
| jrich523 | havent really screwed with that yet, (it will be the next problem) but since im using local VMs i'll need to run a VM as storage | 00:23 |
| Mech422 | sean-k-mooney: and since the cloud stuff is all 'backend' and not revenue generating, I have to make it work with as little as possible | 00:24 |
| sdake | jrich523 note your in for special pain by using virtual machines :) | 00:24 |
| Mech422 | sean-k-mooney: the old systems didn't even have shared storage, so no vm migration at all :-( | 00:24 |
| jrich523 | really all i've done with .21 is setup SSH and some other basic os stuff | 00:24 |
| jrich523 | lol yeah, i miss having corp resources | 00:24 |
| sean-k-mooney | Mech422: ah ok. may i ask which cdn not that i know of many just wondering | 00:24 |
| *** HyperJohnGraham has joined #openstack-kolla | 00:24 | |
| sdake | jrich523 thats good then | 00:24 |
| jrich523 | k | 00:24 |
| sean-k-mooney | Mech422: infact i only know the names of two swiftly and cachefly | 00:25 |
| sdake | jrich523 do you have enough memory in these vms | 00:26 |
| Mech422 | sean-k-mooney: the BIG one is Akami - they're the google of CDN, we're LLNW currently the number 2 I think, then there's Level 3 and Verizon | 00:26 |
| jrich523 | i think i gave them 4gb each, desktop has 32gb so i can up it if need be | 00:26 |
| Mech422 | sean-k-mooney: err...LLNW == Limelight Networks | 00:26 |
| sdake | controller needs 8gb imo | 00:26 |
| sdake | but it may run in 4gb - not certain | 00:26 |
| sdake | compute needs 8gb if you intend to laucnh any type of vm ;) | 00:27 |
| jrich523 | so far its been ok, but.. it managing nothing :) | 00:27 |
| sdake | right its idle | 00:27 |
| sean-k-mooney | Mech422: oh i have herad of limelight networks before and level3 | 00:27 |
| sdake | you will likely get oom errors | 00:27 |
| sdake | and things will break mysteriously | 00:27 |
| sdake | and you will want to hunt us down | 00:27 |
| sean-k-mooney | Mech422: virizon i always associated as an ISP | 00:27 |
| jrich523 | ok, i'll up them shortly | 00:28 |
| Mech422 | sean-k-mooney: yeah - Level 3 is more a tier 1 bandwidth provider, but they started offering CDN as well | 00:28 |
| jrich523 | so, i used the .30, and i got some ssh problems, i'll try that as localhost, see if that resolves that | 00:28 |
| Mech422 | sean-k-mooney: I don't really know what verizon's doing - I'd imagine trying to capture some CDN $$ for shoving bits to their mobile customers ? | 00:29 |
| sean-k-mooney | Mech422: ya maybe. virison has a resonably big fiber install base in ireland/the uk too | 00:30 |
| sean-k-mooney | Mech422: not sure how moch of that market the have states side | 00:31 |
| sean-k-mooney | Mech422: i guess its mostly dominated by twc and comcast | 00:31 |
| *** v1k0d3n has quit IRC | 00:32 | |
| sean-k-mooney | sdake: i have deployed kolla all in one in vms before but i need to give multinode a try | 00:34 |
| Mech422 | sean-k-mooney: oh? huh - in the US they're basically a cell phone provider and thats it... | 00:34 |
| Mech422 | sean-k-mooney: but never really thought of them as a 'real' ISP | 00:34 |
| sean-k-mooney | sdake: i find 8 core 16GB vms with 25-50G of disk quite good for developing | 00:34 |
| Mech422 | sean-k-mooney: jesus - how big is your dev box ? | 00:35 |
| jrich523 | i have 8 cores 4ghz, 32gb ram, about 1.5TB of SSD | 00:36 |
| jrich523 | mostly because i like to play games | 00:36 |
| jrich523 | but the home lab is a bonus :) | 00:36 |
| Mech422 | sean-k-mooney: I run win8 on the desktop, with a 16G KDE based ubuntu VM for work | 00:36 |
| Mech422 | jrich523: heh - I think I have 4 core 3.5GHZ - 32G, and 750G SSD | 00:36 |
| sdake | i have 16 core, 32 thread 256g ram 1.2tb storage on my main workstation | 00:36 |
| sdake | and 8 core 16 thread 32gb ram 400gb storage on my 3 test machines | 00:37 |
| jrich523 | well i guess i have 8 if you count HT | 00:37 |
| sean-k-mooney | sdake: nice | 00:37 |
| jrich523 | 4 cores, 8 logical processors | 00:37 |
| Mech422 | sdake: thats about what my lab nodes are - except for the storage - my lab nodes just have 64(120?) Gig SSDs | 00:38 |
| sean-k-mooney | jrich523: i baught some cheap servers on ebay of very little | 00:38 |
| sdake | and all my storage are intel 750 ssds | 00:38 |
| jrich523 | i considered that, but it eats a lot of power | 00:38 |
| jrich523 | that being said, power here in WA is dirt cheap | 00:38 |
| jrich523 | i found a 960GB SSD for $164 not to long ago, great deal | 00:38 |
| Mech422 | sean-k-mooney: heh - thats where I got mine - 8 node HP S6500 and 2 4 node dell c6100 | 00:38 |
| jrich523 | i could probably "borrow" some from work | 00:38 |
| sbezverk | sean-k-mooney: how I can add br-ex if vswitchd is not running? | 00:38 |
| sean-k-mooney | Mech422: ah the c6100 is nice | 00:39 |
| jrich523 | actually, i could probably setup my own lab at work.. but blah... its a 3 hour flight away... | 00:39 |
| sbezverk | ovsdb-server --run? | 00:39 |
| Mech422 | jrich523: yeah - power is a killer on rack systems - I ended up pulling an extra 40 amps to the 'server room' | 00:39 |
| sean-k-mooney | sbezverk: sudo ovs-vsctl --no-wait add-br br-ex | 00:39 |
| jrich523 | oh jrdud | 00:39 |
| jrich523 | jesus | 00:39 |
| Mech422 | sean-k-mooney: yeah - I actually like the HPs more though. the c6100 ipmi is flakey as all shit | 00:39 |
| jrich523 | lol did you steal that from your stove/range? :) | 00:40 |
| sean-k-mooney | Mech422: i have a hp 2xsl170z | 00:40 |
| Mech422 | jrich523: nah - my buddy is an electrician, so he ran a bunch of 20 amp circuits around for me | 00:40 |
| jrich523 | oh nice | 00:40 |
| Mech422 | sean-k-mooney: oh? how you like it? | 00:40 |
| sean-k-mooney | Mech422: its basically the hp version of the c6100. i only use it for a home lab but it works well. | 00:41 |
| Mech422 | jrich523: sounds is the other big problem - the servers are LOUD | 00:41 |
| jrich523 | i'd be comfortable running that myself... just... dont want to pay for the cost of running gear i wont really be using | 00:41 |
| *** pbourke has joined #openstack-kolla | 00:41 | |
| jrich523 | yeah, doesnt seem like much in a server room | 00:41 |
| jrich523 | until you get it home :) | 00:41 |
| sean-k-mooney | Mech422: the c6100 has better storage having 3 3.5" bays per server vs 1 | 00:41 |
| jrich523 | back in my younger days i use to do that stuff at home | 00:41 |
| jrich523 | now my pc is powerful enough to do what i need | 00:42 |
| Mech422 | sean-k-mooney: yeah - but I don't really need bulk storage - these nodes get torn down 2-3xday to install kolla and whatnot | 00:42 |
| Mech422 | sean-k-mooney: I didn't even bother with mdraid - they only have a single SSD each | 00:42 |
| sdake | jrich523 i am quickly runing out of steam | 00:42 |
| jrich523 | no worries | 00:43 |
| jrich523 | dont stay here on my account | 00:43 |
| sdake | jrich523 got anything to work with in terms of an invetory file to paste | 00:43 |
| sdake | jrich523 i finish the job | 00:43 |
| jrich523 | just about to run it | 00:43 |
| sean-k-mooney | Mech422: yep. i have a dedicated ceph node that i build from an old pc for home storage and just throun 120g ssd in the servers for local storage | 00:43 |
| sdake | paste it first pls | 00:43 |
| jrich523 | last attempt failed because i didnt have docker on 21 | 00:43 |
| jrich523 | just changed the .30 to localhost | 00:43 |
| jrich523 | to see if that avoids the ssh problem | 00:43 |
| Mech422 | sean-k-mooney: oh - you mean the c6100 has better storage then the HP ? | 00:43 |
| jrich523 | and the 21 deployment failed because no docker | 00:43 |
| sdake | paste inventory pls :) | 00:43 |
| sdake | because i am going to bed shortly | 00:44 |
| sean-k-mooney | Mech422: yes | 00:44 |
| jrich523 | http://paste.openstack.org/show/561555/ | 00:44 |
| sdake | ok you can't have localhost in there | 00:44 |
| sdake | you ahve to have the ip address of the control node | 00:44 |
| jrich523 | yeah i'll have to setup the ssh key | 00:44 |
| sdake | its more fundamental then that | 00:45 |
| jrich523 | but it seems to be deploying to 21 just fine | 00:45 |
| sdake | that address is used all ove rthe place | 00:45 |
| sdake | yes - pretty ure compute wont be able to communicate with it | 00:45 |
| jrich523 | yeah only 2 systems right now | 00:45 |
| sdake | rather 30 | 00:45 |
| jrich523 | fatal: [192.168.11.21]: FAILED! => {"changed": false, "failed": true, "module_stderr": "", "module_stdout": "Traceback (most recent call last):\r\n File \"/tmp/ansible_rRtr1Y/ansible_module_kolla_docker.py\", line 194, in <module>\r\n import docker\r\nImportError: No module named docker\r\n", "msg": "MODULE FAILURE", "parsed": false} | 00:45 |
| sdake | the compute and control wont be able to contact each other with that inventory | 00:45 |
| jrich523 | thats the only 'tricky' error im getting | 00:46 |
| sdake | sudo pip install docker-py | 00:46 |
| Mech422 | sean-k-mooney: wow - those HP setups are about half the price/node of the C6100 | 00:46 |
| jrich523 | k | 00:46 |
| sdake | sudo pip install docker-py==1.6.0 should do | 00:46 |
| sean-k-mooney | Mech422: that reminds me i need to check out kolla external ceph support. | 00:46 |
| jrich523 | i'll do that and fix the ssh key stuff, that will be enough for today :) | 00:46 |
| sdake | enoy i am off | 00:46 |
| sdake | later folks | 00:46 |
| sean-k-mooney | Mech422: ya they are quite old but cheap and fine for openstack dev | 00:46 |
| jrich523 | thanks sdake, have a good one | 00:46 |
| *** sdake has quit IRC | 00:46 | |
| Mech422 | sean-k-mooney: yeah, I've been pretty happy with the old gear - especially having ipmi | 00:47 |
| Mech422 | sean-k-mooney: Foreman is great for provisioning stuff too - makes it pretty painless to re-install a clean image and test stuff | 00:47 |
| jrich523 | i really hate HP, partly because their iLo requires an enterprise lic to be useful | 00:47 |
| sean-k-mooney | Mech422: i paid about £350-400 for mine on ebay last chirstmas. i have been using them to do the bifros testing. | 00:48 |
| Mech422 | jrich523: yeah - the fancy ilo features would be nice - some of the used machines come with it. Mine don't - but ipmitool works well on them | 00:48 |
| jrich523 | yeah, i guess its not generally needed these days... but... would still be nice | 00:48 |
| Mech422 | I paid about $2k US for 4 node c6100 - and $2k US for 8 node HP with similiar specs | 00:49 |
| jrich523 | not too bad | 00:49 |
| Mech422 | yeah - the used stuff is <= $250/node now :-) pretty sweet! | 00:49 |
| jrich523 | lol i should create an alias for yup=yum for how often i misstype that | 00:49 |
| Mech422 | anyway - I'm off - have a great weekend everyone ! | 00:50 |
| jrich523 | cya Mech422 | 00:51 |
| sean-k-mooney | Mech422: http://www.ebay.co.uk/itm/Dell-PowerEdge-C6100-VMware-Cloud-Node-Server-8x-E5506-32-Cores-48GB-RAM-2U-/141904567567?hash=item210a2bdd0f | 00:51 |
| jrich523 | damn | 00:52 |
| sean-k-mooney | Mech422: they are selling full racks of them currently | 00:52 |
| jrich523 | i got a pc for my home automation/security system... and i just grabbed a Dell optiplex 3020, i paid about the same amoung | 00:53 |
| sean-k-mooney | 15 for only 7.5K http://www.ebay.co.uk/itm/Dell-PowerEdge-Full-Server-Rack-15x-C6100-120-Quad-Core-CPUs-2880GB-DDR3-RAM-/142028516759?hash=item21118f2d97 | 00:53 |
| jrich523 | no drives | 00:53 |
| sean-k-mooney | i really would nto want to pay the power bill though unless i was makeing money from them | 00:53 |
| jrich523 | and no caddies! | 00:54 |
| jrich523 | geesh | 00:54 |
| jrich523 | lol if i got that im sure i could host it at work | 00:54 |
| jrich523 | we have a rather large integration center i could just toss it in the corner | 00:54 |
| sean-k-mooney | jrich523: ya im realy wanting to get our new kolla cluster working at work | 00:55 |
| jrich523 | whats wrong with it? | 00:58 |
| jrich523 | the quick start guide not working out for you? :) | 00:58 |
| sean-k-mooney | waithing for harddrive to arrive | 00:59 |
| jrich523 | ahhh, yeah those are kinda important | 00:59 |
| jrich523 | we have openstack at work | 00:59 |
| jrich523 | we sell metapod | 00:59 |
| jrich523 | which is partly why im learning openstack | 00:59 |
| jrich523 | but god does metapod blow | 00:59 |
| sean-k-mooney | i work at intel and we can get it provided openstack acess but out team is dedicated to openatck development so we have been trying to get our own clould inplace for a while | 01:00 |
| jrich523 | right now, we have 5 in the setup, with like 5 vms, that are idle, and if i go to create one it cant 'schedule' it | 01:00 |
| jrich523 | oh nice | 01:00 |
| jrich523 | yeah i ust want to learn the ins and outs and what have yous | 01:00 |
| sean-k-mooney | we just got kolla with ironic as a service working during the week | 01:01 |
| jrich523 | nice | 01:01 |
| jrich523 | im really liking how kolla does things... | 01:01 |
| jrich523 | but you know, we cant make money off it... so.. cisco it is! | 01:01 |
| sean-k-mooney | we are going to use that to pull our 40 nodes into a baremetal clound then deploy 2 vmclounds ontop of that | 01:01 |
| sean-k-mooney | 1 for ci the other for dev use | 01:02 |
| jrich523 | cloudception :) | 01:02 |
| sean-k-mooney | well if it means i dont have to go donw to the lab and manually reinstall oss then yes. its kina like triple0 without the puppet | 01:03 |
| jrich523 | all right, its time to grab some dinner | 01:03 |
| jrich523 | you have yourself a good evening | 01:03 |
| sean-k-mooney | ya i should be heading off too. enjoy the weekend everyone | 01:03 |
| *** salv-orlando has joined #openstack-kolla | 01:04 | |
| *** sean-k-mooney is now known as sean-k-mooneyAFK | 01:04 | |
| *** kbaegis has quit IRC | 01:09 | |
| *** HyperJohnGraham has quit IRC | 01:12 | |
| *** aNupoisc has quit IRC | 01:14 | |
| *** salv-orlando has quit IRC | 01:15 | |
| *** daneyon has joined #openstack-kolla | 01:49 | |
| *** daneyon has quit IRC | 01:54 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Create the nova instances directory when need during starting https://review.openstack.org/357898 | 02:02 |
| *** gfhellma has quit IRC | 02:07 | |
| openstackgerrit | Shaun Smekel proposed openstack/kolla: Add dockerfiles for keystone fernet https://review.openstack.org/351139 | 02:08 |
| openstackgerrit | Shaun Smekel proposed openstack/kolla: Add dockerfiles for keystone fernet https://review.openstack.org/351139 | 02:08 |
| *** salv-orlando has joined #openstack-kolla | 02:15 | |
| *** salv-orlando has quit IRC | 02:17 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: DO_NOT_MERGE: test deploy ubuntu source https://review.openstack.org/358157 | 02:31 |
| *** v1k0d3n has joined #openstack-kolla | 02:32 | |
| openstackgerrit | Shaun Smekel proposed openstack/kolla: Add dockerfiles for keystone fernet https://review.openstack.org/351139 | 02:34 |
| *** v1k0d3n has quit IRC | 02:37 | |
| *** dave-mccowan has joined #openstack-kolla | 02:43 | |
| *** HyperJohnGraham has joined #openstack-kolla | 02:43 | |
| openstackgerrit | Shaun Smekel proposed openstack/kolla: Customizations for MariaDB https://review.openstack.org/354422 | 02:56 |
| openstackgerrit | Shaun Smekel proposed openstack/kolla: Customizations for MariaDB https://review.openstack.org/354422 | 02:58 |
| openstackgerrit | Shaun Smekel proposed openstack/kolla: Customizations for RabbitMQ https://review.openstack.org/354409 | 03:00 |
| openstackgerrit | Shaun Smekel proposed openstack/kolla: Customizations for Ironic https://review.openstack.org/354427 | 03:02 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: DO_NOT_MERGE: test deploy ubuntu source https://review.openstack.org/358157 | 03:15 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Remove the duplicated RUN directive in the elasticsearch Dockerfile https://review.openstack.org/358161 | 03:21 |
| *** salv-orlando has joined #openstack-kolla | 03:21 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Create the nova instances directory when need during starting https://review.openstack.org/357898 | 03:22 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: DO_NOT_MERGE: test deploy ubuntu source https://review.openstack.org/358157 | 03:22 |
| *** salv-orlando has quit IRC | 03:32 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Remove the duplicated RUN directive in the elasticsearch Dockerfile https://review.openstack.org/358161 | 03:39 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Create the nova instances directory when need during starting https://review.openstack.org/357898 | 03:41 |
| *** unicell1 has joined #openstack-kolla | 03:48 | |
| *** ad_rien_ has joined #openstack-kolla | 03:50 | |
| *** unicell has quit IRC | 03:50 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: DO_NOT_MERGE: test deploy ubuntu source https://review.openstack.org/358157 | 03:50 |
| *** dave-mccowan has quit IRC | 03:57 | |
| *** harlowja_at_home has joined #openstack-kolla | 04:02 | |
| *** zhurong has joined #openstack-kolla | 04:06 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Remove the RUN directive for macros.install_packages in the Dockerfile https://review.openstack.org/358161 | 04:29 |
| *** harlowja_at_home has quit IRC | 04:31 | |
| *** daneyon has joined #openstack-kolla | 04:32 | |
| *** salv-orlando has joined #openstack-kolla | 04:34 | |
| openstackgerrit | Shaun Smekel proposed openstack/kolla: Urgent: Fixes build failures https://review.openstack.org/358164 | 04:35 |
| *** daneyon has quit IRC | 04:36 | |
| *** salv-orlando has quit IRC | 04:37 | |
| *** zhurong has quit IRC | 04:39 | |
| Jeffrey4l_ | smekel_, ping | 04:43 |
| Jeffrey4l_ | smekel_, https://review.openstack.org/#/c/358161/3/docker/congress/congress-base/Dockerfile.j2 | 04:43 |
| patchbot | Jeffrey4l_: patch 358161 - kolla - Remove the RUN directive for macros.install_packag... | 04:43 |
| Jeffrey4l_ | smekel_, i do not touch anything. it is my vim editor add a '\n' char at the last lines. | 04:44 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: DO_NOT_MERGE: test deploy ubuntu source https://review.openstack.org/358157 | 04:48 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Fix the deploy Ubuntu source gate https://review.openstack.org/358157 | 05:36 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Remove the RUN directive for macros.install_packages in the Dockerfile https://review.openstack.org/358161 | 05:36 |
| *** salv-orlando has joined #openstack-kolla | 05:46 | |
| *** salv-orlando has quit IRC | 05:57 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Pin to the Ubuntu 16.04 https://review.openstack.org/329454 | 06:27 |
| *** ad_rien_ has joined #openstack-kolla | 06:30 | |
| *** ad_rien_ has quit IRC | 06:31 | |
| *** v1k0d3n has joined #openstack-kolla | 06:34 | |
| *** v1k0d3n has quit IRC | 06:39 | |
| *** sbezverk_ has joined #openstack-kolla | 06:46 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Fix the deploy Ubuntu source gate https://review.openstack.org/358157 | 06:47 |
| *** papacz1 has joined #openstack-kolla | 06:48 | |
| *** unicell1 has quit IRC | 06:48 | |
| *** peterjenkins_ has joined #openstack-kolla | 06:49 | |
| *** mdnadeem has joined #openstack-kolla | 06:49 | |
| *** Daviey_ has joined #openstack-kolla | 06:50 | |
| *** akwasnie__ has joined #openstack-kolla | 06:50 | |
| *** DuncanT_ has joined #openstack-kolla | 06:50 | |
| *** mgkwill_ has joined #openstack-kolla | 06:52 | |
| *** dgonzalez_ has joined #openstack-kolla | 06:52 | |
| openstackgerrit | Christian Berendt proposed openstack/kolla: Add cloudkitty containers for type source https://review.openstack.org/349295 | 06:54 |
| *** peterjenkins has quit IRC | 06:54 | |
| *** mdnadeem_ has quit IRC | 06:54 | |
| *** dgonzalez has quit IRC | 06:54 | |
| *** papacz has quit IRC | 06:54 | |
| *** akwasnie has quit IRC | 06:54 | |
| *** mgkwill has quit IRC | 06:54 | |
| *** sbezverk has quit IRC | 06:54 | |
| *** DuncanT has quit IRC | 06:54 | |
| *** jascott1 has quit IRC | 06:54 | |
| *** kklimonda has quit IRC | 06:54 | |
| *** Daviey has quit IRC | 06:54 | |
| *** dgonzalez_ is now known as dgonzalez | 06:54 | |
| *** peterjenkins_ is now known as peterjenkins | 06:54 | |
| *** akwasnie__ is now known as akwasnie | 06:55 | |
| *** jascott1 has joined #openstack-kolla | 06:55 | |
| *** mgkwill_ is now known as mgkwill | 06:56 | |
| *** salv-orlando has joined #openstack-kolla | 06:59 | |
| *** DuncanT_ is now known as DuncanT | 07:01 | |
| *** salv-orlando has quit IRC | 07:02 | |
| *** daneyon has joined #openstack-kolla | 07:14 | |
| *** daneyon has quit IRC | 07:19 | |
| *** fragatina has quit IRC | 07:46 | |
| *** salv-orlando has joined #openstack-kolla | 08:03 | |
| *** ad_rien_ has joined #openstack-kolla | 08:09 | |
| *** salv-orlando has quit IRC | 08:16 | |
| *** zhurong has joined #openstack-kolla | 08:18 | |
| *** ad_rien_ has quit IRC | 08:27 | |
| *** dwalsh has joined #openstack-kolla | 08:37 | |
| *** ad_rien_ has joined #openstack-kolla | 08:49 | |
| *** kklimonda has joined #openstack-kolla | 08:56 | |
| *** daneyon has joined #openstack-kolla | 09:02 | |
| *** daneyon has quit IRC | 09:07 | |
| *** ad_rien_ has quit IRC | 09:09 | |
| *** salv-orlando has joined #openstack-kolla | 09:16 | |
| *** salv-orlando has quit IRC | 09:23 | |
| *** ad_rien_ has joined #openstack-kolla | 09:33 | |
| *** ad_rien_ has quit IRC | 09:42 | |
| *** ad_rien_ has joined #openstack-kolla | 09:43 | |
| *** ad_rien_ has quit IRC | 09:45 | |
| *** sdake has joined #openstack-kolla | 09:51 | |
| *** aswadr_ has joined #openstack-kolla | 09:52 | |
| *** ad_rien_ has joined #openstack-kolla | 09:52 | |
| *** ad_rien_ has quit IRC | 09:54 | |
| *** zhurong has quit IRC | 09:55 | |
| *** salv-orlando has joined #openstack-kolla | 10:28 | |
| *** matrohon has joined #openstack-kolla | 10:31 | |
| *** salv-orlando has quit IRC | 10:31 | |
| sdake | morning | 10:31 |
| *** sdake has quit IRC | 10:32 | |
| *** sdake has joined #openstack-kolla | 10:34 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Fix the deploy Ubuntu source gate https://review.openstack.org/358157 | 10:35 |
| *** v1k0d3n has joined #openstack-kolla | 10:36 | |
| *** v1k0d3n has quit IRC | 10:40 | |
| *** egonzalez90 has joined #openstack-kolla | 10:43 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Pin to the Ubuntu 16.04 https://review.openstack.org/329454 | 10:47 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: DO NOT MERGE: TEST MITAKA BRANCH https://review.openstack.org/299182 | 10:48 |
| *** daneyon has joined #openstack-kolla | 10:51 | |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla: Remove unnecessary host at haproxy in site.yml https://review.openstack.org/355861 | 10:53 |
| *** daneyon has quit IRC | 10:55 | |
| *** dwalsh has quit IRC | 11:06 | |
| sdake | egonzalez90 ping | 11:08 |
| egonzalez90 | morning sdake | 11:09 |
| sdake | egonzalez90 sup - welcome to the core reviewer team! | 11:09 |
| sdake | not sure if you cuaght the email | 11:09 |
| egonzalez90 | thanks o/ | 11:09 |
| sdake | quick primer | 11:09 |
| sdake | you can +2 or -2 any change now | 11:09 |
| sdake | we do not approve our own changes | 11:09 |
| sdake | if we are a co-author on a change, we may not +2 it | 11:10 |
| sdake | before workflow can be applied, 2 +2 votes are required in total | 11:10 |
| sdake | workflow button merges the patch | 11:10 |
| sdake | its important that the gates that are working remain green during the review | 11:10 |
| sdake | if they aren't green, use 'recheck" in the post comment field | 11:10 |
| sdake | this will trigger a revaliation via ci | 11:11 |
| egonzalez90 | yup, we talked at the last meeting about that | 11:11 |
| sdake | atm the ubuntu deploy gate is broken and the bandit gate is broken | 11:12 |
| sdake | bandit gate should be fixed soon | 11:12 |
| sdake | i'm going to look at the ubuntu gate now | 11:12 |
| sdake | any quesstions? | 11:13 |
| egonzalez90 | not atm, but for sure i'll have | 11:15 |
| *** matrohon has quit IRC | 11:22 | |
| openstackgerrit | Merged openstack/kolla: Remove the RUN directive for macros.install_packages in the Dockerfile https://review.openstack.org/358161 | 11:26 |
| *** bootsha has joined #openstack-kolla | 11:29 | |
| *** ad_rien_ has joined #openstack-kolla | 11:31 | |
| *** bootsha has quit IRC | 11:34 | |
| *** salv-orlando has joined #openstack-kolla | 11:38 | |
| *** salv-orlando has quit IRC | 11:41 | |
| *** ad_rien_ has quit IRC | 11:47 | |
| *** kbaegis has joined #openstack-kolla | 11:49 | |
| *** dwalsh has joined #openstack-kolla | 11:51 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Pin to the Ubuntu 16.04 https://review.openstack.org/329454 | 11:52 |
| *** egonzalez90 has quit IRC | 11:59 | |
| *** sdake has quit IRC | 12:06 | |
| *** sdake has joined #openstack-kolla | 12:06 | |
| *** JoseMello has joined #openstack-kolla | 12:07 | |
| *** dwalsh has quit IRC | 12:12 | |
| *** egonzalez90 has joined #openstack-kolla | 12:13 | |
| *** egonzalez90 is now known as Guest66754 | 12:13 | |
| kbaegis | Do you guys need a architecture diagram? | 12:15 |
| *** Guest66754 has quit IRC | 12:16 | |
| *** egonzalez90_ has joined #openstack-kolla | 12:16 | |
| openstackgerrit | Christian Berendt proposed openstack/kolla: Customizations for Neutron https://review.openstack.org/350777 | 12:24 |
| *** salv-orlando has joined #openstack-kolla | 12:42 | |
| *** salv-orlando has quit IRC | 12:47 | |
| *** dave-mccowan has joined #openstack-kolla | 12:51 | |
| *** ad_rien_ has joined #openstack-kolla | 12:58 | |
| *** ad_rien_ has quit IRC | 13:12 | |
| sdake | Jeffrey4l_ I was looking to fix the gate | 13:19 |
| sdake | Jeffrey4l_ I noticed maybe you already have | 13:19 |
| sdake | can you explain this change: https://review.openstack.org/#/c/329454/15/docker/kolla-toolbox/Dockerfile.j2 | 13:19 |
| patchbot | sdake: patch 329454 - kolla - Pin to the Ubuntu 16.04 | 13:19 |
| *** aswadr_ has quit IRC | 13:21 | |
| Jeffrey4l_ | sdake, yes. it is. | 13:21 |
| sdake | can ou explain the --no-binary | 13:23 |
| Jeffrey4l_ | sdake, the root cause is: ci provide pip mysql-python wheel package, which contains pre-build _mysql.so | 13:23 |
| sdake | i wuld like ot break that change out separately from pinning | 13:23 |
| sdake | how did you figur ethat out | 13:23 |
| Jeffrey4l_ | _mysql.so in the wheel depends on libmysql.so.20 | 13:23 |
| Jeffrey4l_ | but in the ubuntu, it only has libmysq.so.18 | 13:23 |
| sdake | this has been a super long standing problem | 13:23 |
| sdake | --no-binary does what? | 13:24 |
| Jeffrey4l_ | i disable the wheel index in the ci by using this https://review.openstack.org/358157 | 13:24 |
| Jeffrey4l_ | sdake, --no-binary mysql-python stop using the pre-build wheel package and install from source | 13:25 |
| sdake | Jeffrey4l_ - I dont think we want to disable to prebuilt wheel package | 13:25 |
| sdake | it will i think negatively impact gating performance | 13:25 |
| sdake | package/packages | 13:25 |
| sdake | but --no-binary wfm | 13:25 |
| Jeffrey4l_ | sdake, we shouldn't use the pre-build package. | 13:26 |
| *** bootsha has joined #openstack-kolla | 13:26 | |
| Jeffrey4l_ | i have no idea where the wheel is come from. | 13:26 |
| sdake | https://bugs.launchpad.net/kolla/+bug/1615231 | 13:27 |
| openstack | Launchpad bug 1615231 in kolla "ubuntu kolla-toolbox gate fails with module mysqldb is required" [Critical,Confirmed] - Assigned to Steven Dake (sdake) | 13:27 |
| sdake | can you breka out the no-binary into that bug | 13:27 |
| sdake | or i can do it | 13:27 |
| Jeffrey4l_ | without the wheel index. we can test the dependency is OK in the images. | 13:27 |
| sdake | yup | 13:27 |
| sdake | i hear what your syaing | 13:27 |
| Jeffrey4l_ | yep. | 13:27 |
| sdake | it will give us more reliable gating results | 13:27 |
| sdake | i am just concenred about gate performance | 13:28 |
| Jeffrey4l_ | So i prefer to disable wheel package. | 13:28 |
| sdake | can you run a comparison of one versus the other? | 13:28 |
| Jeffrey4l_ | sdake, np | 13:28 |
| sdake | if its not too much impct then wfm | 13:28 |
| sdake | if its like DOUBLE thats a problem | 13:28 |
| Jeffrey4l_ | sdake, OK | 13:28 |
| sdake | ok to which requet? :) | 13:29 |
| sdake | i can break out the --no-binary into that bug if you like | 13:29 |
| Jeffrey4l_ | if we disable the wheel finally, no need add the no-binary | 13:29 |
| Jeffrey4l_ | just do not add | 13:30 |
| Jeffrey4l_ | [global] | 13:30 |
| Jeffrey4l_ | wheel-index = xxx | 13:30 |
| Jeffrey4l_ | will work | 13:30 |
| Jeffrey4l_ | like this PS does sdake | 13:30 |
| sdake | moment to think please | 13:30 |
| sdake | I wonder if someone would turn on a wheel cache outside kolla (not in the gate) | 13:31 |
| sdake | but just during their normal ci workflow of kolla as a product | 13:31 |
| sdake | is that a reasonable thing to do, or would nobody do that? | 13:32 |
| Jeffrey4l_ | sdake, it is possible. | 13:32 |
| sdake | possible and reasonable mean two different things :) | 13:32 |
| sdake | i alreay know its possible - we do it in the gate | 13:33 |
| Jeffrey4l_ | possible and reasonable. | 13:33 |
| sdake | ok - then --no-binary should probably stay | 13:33 |
| Jeffrey4l_ | But it is his work to ensure the wheel work for ubuntu xenial | 13:33 |
| sdake | we are talking about two thingsh ere | 13:33 |
| sdake | 1. --no-binary | 13:34 |
| Jeffrey4l_ | sdake, no. if we want to use --no-binary in kolla. we should add it into all the package which need build. | 13:34 |
| sdake | 2. wheel cche removal from oour ci | 13:34 |
| sdake | Jeffrey4l_ yup that thought just crossed my mind | 13:34 |
| Jeffrey4l_ | in that way, the wheel will never work. | 13:34 |
| sdake | i see so no point in a wheel cache in that scenario | 13:34 |
| sdake | what does --no-binary do? | 13:35 |
| Jeffrey4l_ | the wheel package is just like what the ubuntu repo source. | 13:35 |
| Jeffrey4l_ | the end-user should ensure the repo source/wheel works on ubuntu xenial | 13:35 |
| Jeffrey4l_ | sdake, --no-binary just disable one/multi package to stop use wheel. | 13:36 |
| Jeffrey4l_ | packages. | 13:36 |
| Jeffrey4l_ | sdake, pip install --help show the usage. | 13:37 |
| sdake | does it do that - or does it tell pip to rebuild anything with a .so in it? | 13:37 |
| openstackgerrit | Shaun Smekel proposed openstack/kolla: Add dockerfiles for keystone fernet https://review.openstack.org/351139 | 13:37 |
| Jeffrey4l_ | sdake, do not use wheel package. just build from the source code. | 13:37 |
| Jeffrey4l_ | for the mysql-python, it will build its own _mysql.so package. | 13:38 |
| Jeffrey4l_ | compare the PS6 and PS7 ubuntu build time in this https://review.openstack.org/#/c/358157/ | 13:38 |
| patchbot | Jeffrey4l_: patch 358157 - kolla - Fix the deploy Ubuntu source gate | 13:38 |
| Jeffrey4l_ | PS6 use the wheel and PS disable the wheel. | 13:38 |
| Jeffrey4l_ | the time is almost the same. | 13:38 |
| Jeffrey4l_ | it do not take much time to build the pip package. 1. there are few package need build 2. much time is used for download file, no the cpu time. | 13:39 |
| Jeffrey4l_ | sdake, ^ | 13:40 |
| sdake | ps5 and ps6 look exactly the same | 13:40 |
| sdake | Jeffrey4l_ thanks I read the help - i undderstand what it does now | 13:41 |
| *** dmueller-evoila has joined #openstack-kolla | 13:41 | |
| sdake | ok 6 and 7 moment | 13:41 |
| Jeffrey4l_ | yes | 13:41 |
| sdake | ps7 is red | 13:42 |
| Jeffrey4l_ | hmm sorry, | 13:42 |
| Jeffrey4l_ | should be PS6 vs PS8 | 13:42 |
| Jeffrey4l_ | sdake, ^ | 13:42 |
| sdake | build centos7 wheel - 17 min build centos7 no wheeel = 38min | 13:44 |
| Jeffrey4l_ | sdake, centos7 using extra repo. | 13:45 |
| sdake | expand? | 13:45 |
| sdake | which repo in particular | 13:45 |
| Jeffrey4l_ | centos7 is not using CI repo. So the network is the big issue. | 13:45 |
| Jeffrey4l_ | sdake, all the centos repo. | 13:46 |
| sdake | centos7 should be using the wheel mirror | 13:46 |
| sdake | atlaet last time i looked at teh code | 13:46 |
| sdake | should as in currently is doing so | 13:46 |
| Jeffrey4l_ | sdake, just compare the ubuntu source build . Because it using the repo from CI. | 13:47 |
| sdake | right i saw that | 13:47 |
| sdake | its not using all ubuntu mirrors but close to it | 13:47 |
| Jeffrey4l_ | sdake, but centos7 using repo not in CI. | 13:47 |
| sdake | so why the big delta between ps6 and ps8 on centos source? | 13:48 |
| Jeffrey4l_ | So the network is the big issue for the build time. | 13:48 |
| sdake | i see- so you think its unrelated to wheel | 13:48 |
| Jeffrey4l_ | sdake, in default centos using repo mirror. no body know which mirror site he will use. | 13:48 |
| Jeffrey4l_ | yes. | 13:48 |
| sdake | you think its related to all the million packages pulled down rom yum repos | 13:48 |
| sdake | got it | 13:49 |
| sdake | ok wfm | 13:49 |
| Jeffrey4l_ | right. | 13:49 |
| Jeffrey4l_ | So https://review.openstack.org/358157 is really for review/merge. | 13:49 |
| Jeffrey4l_ | then i will remove the --no-binary from https://review.openstack.org/329454 | 13:50 |
| sdake | Jeffrey4l_ - thanks dude - i was going to sort that out this mroning but looks like your on it :) | 13:51 |
| Jeffrey4l_ | ya. ps is coming :) | 13:51 |
| sdake | so 358157 ready for review you mean? | 13:52 |
| *** salv-orlando has joined #openstack-kolla | 13:52 | |
| Jeffrey4l_ | sdake, yep | 13:52 |
| Jeffrey4l_ | 329454 will rebase on it. | 13:53 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Pin to the Ubuntu 16.04 https://review.openstack.org/329454 | 13:55 |
| sdake | Jeffrey4l_ couple small changes needed in your 358157 | 13:57 |
| Jeffrey4l_ | sdake, OK. thanks | 13:57 |
| *** salv-orlando has quit IRC | 13:59 | |
| *** bootsha has quit IRC | 14:00 | |
| sdake | Jeffrey4l_ nice work on sorting out the gate problem - i think i'm going to take on this bandit patch of pauls to fix it up | 14:00 |
| sdake | so our gate is perffectly green | 14:00 |
| Jeffrey4l_ | yep. the all green ;p | 14:00 |
| openstackgerrit | Shaun Smekel proposed openstack/kolla: Fix incorrectly implemented footer blocks https://review.openstack.org/358193 | 14:01 |
| openstackgerrit | Shaun Smekel proposed openstack/kolla: Customizations for Ironic https://review.openstack.org/354427 | 14:03 |
| *** pbourke has quit IRC | 14:13 | |
| sdake | yay I upgraded to office 2016 | 14:15 |
| sdake | and now have no calendar!! | 14:15 |
| * sdake facepalms | 14:15 | |
| sdake | http://knowyourmeme.com/memes/facepalm | 14:16 |
| sdake | picard does the best facepalm | 14:16 |
| *** ad_rien_ has joined #openstack-kolla | 14:17 | |
| *** ad_rien_ has quit IRC | 14:20 | |
| sdake | Jeffrey4l_ can you resolve https://review.openstack.org/#/c/358157/ comments so I can +2 before you leave for the day | 14:23 |
| patchbot | sdake: patch 358157 - kolla - Fix the deploy Ubuntu source gate | 14:23 |
| *** egonzalez90_ has quit IRC | 14:23 | |
| Jeffrey4l_ | sdake, working on it. | 14:23 |
| Jeffrey4l_ | moment. | 14:23 |
| sdake | cool | 14:23 |
| sdake | sorry dont mean to micromanage | 14:23 |
| sdake | i'd just like the gate green heading into the weekend | 14:24 |
| *** salv-orlando has joined #openstack-kolla | 14:25 | |
| *** ad_rien_ has joined #openstack-kolla | 14:27 | |
| *** sdake_ has joined #openstack-kolla | 14:27 | |
| *** ad_rien_ has quit IRC | 14:29 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Fix Ubuntu source gate deployment https://review.openstack.org/358157 | 14:29 |
| Jeffrey4l_ | sdake, ^^ | 14:29 |
| *** sdake has quit IRC | 14:29 | |
| *** ad_rien_ has joined #openstack-kolla | 14:31 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Pin to Ubuntu 16.04 https://review.openstack.org/329454 | 14:31 |
| jrich523 | Unknown error message: Tag 3.0.0 not found in repository docker.io/kolla/centos-binary-kolla-toolbox"} | 14:36 |
| *** ad_rien_ has quit IRC | 14:36 | |
| jrich523 | ugh i had this before, but i forget how to resolve it | 14:36 |
| jrich523 | sdake_: you around? | 14:37 |
| sdake_ | jrich523 as always | 14:38 |
| jrich523 | feeling any better? | 14:38 |
| sdake_ | yup | 14:38 |
| sdake_ | new day | 14:38 |
| jrich523 | super :) | 14:39 |
| *** hieulq_ has joined #openstack-kolla | 14:41 | |
| jrich523 | how do i resolve the docker tag? i thought it was something in globals, but, that hasnt changed | 14:42 |
| *** inc0 has joined #openstack-kolla | 14:46 | |
| inc0 | good morning | 14:47 |
| jrich523 | morning inc0 | 14:47 |
| sdake_ | sup inc0 | 14:47 |
| sdake_ | Jeffrey4l_ has mostlly go tthe gate fixedd | 14:48 |
| inc0 | just dumped our rally results from ceph installation | 14:48 |
| sdake_ | jrich523 the docker tagcomes from openstack_release | 14:48 |
| inc0 | bunch of fails, gonna analyze | 14:48 |
| inc0 | also broke my headphones. | 14:48 |
| jrich523 | that sucks, what sort of headphones are you rocking? | 14:49 |
| inc0 | fuck. But at least now I have reason to buy fancier ones | 14:49 |
| jrich523 | lol | 14:49 |
| inc0 | ones I had were superlux hmc631 | 14:49 |
| inc0 | and I really liked them, had them for 3 years | 14:49 |
| sdake_ | i'm somewhat fond of the the B&W p7 - replaceable cable | 14:49 |
| inc0 | but cable broke, this is how I break headphones | 14:49 |
| jrich523 | ahh bummer | 14:50 |
| jrich523 | i got some audio technics (ATH-M50x) that i slap a mic on | 14:50 |
| sdake_ | i have b&w p7s and audeze lcd4 headphones | 14:51 |
| inc0 | I thought of ATH when I was buying these, are you happy with it? | 14:51 |
| jrich523 | very | 14:51 |
| jrich523 | and i've gone through a lot | 14:51 |
| jrich523 | im also a fan of the sony 700 series, but they are a tad heavy | 14:51 |
| jrich523 | i think these are my fav based on the 5 or so pairs i've gotten | 14:52 |
| sdake_ | lcd4's rock but they are not at all portable | 14:52 |
| sdake_ | hence the need for the b&w | 14:52 |
| jrich523 | also the sennheiser hd558 is surprisingly good for the price | 14:52 |
| jrich523 | altho not foldable, so... portability is meh | 14:53 |
| jrich523 | http://v-moda.com/boompro-microphone/ | 14:53 |
| jrich523 | then get a 2.5 to 3.5 adapater and that mic | 14:53 |
| jrich523 | works well | 14:53 |
| jrich523 | the ATH is still a good price too, right around 120 | 14:55 |
| inc0 | I don't care about foldable, but actually pluggable cable would be a win | 14:55 |
| inc0 | every headphones I broke (quite a few) were due to cable breakage | 14:55 |
| jrich523 | thats sort of a must | 14:55 |
| inc0 | and also did a lot of soldiering of these | 14:55 |
| jrich523 | lol yeah thats never fun | 14:56 |
| jrich523 | i did have to take these apart to remove the.. umm... pin | 14:56 |
| inc0 | soldiering is fun, but there is only so much you can do | 14:56 |
| jrich523 | but its easy | 14:56 |
| sdake_ | b&w p7s - only one i know of with removable cable | 14:56 |
| inc0 | also I didn't bring my soldiering rig to US | 14:56 |
| sdake_ | and also foldable | 14:56 |
| inc0 | but a bit too pricey | 14:57 |
| inc0 | I'd look for something ~200$ | 14:57 |
| sdake_ | b&w is ~250 right? | 14:57 |
| jrich523 | i'd highly recommend the AT's, most pairs i get are right around $150, and these ahve been the best | 14:58 |
| sdake_ | oh 400 | 14:58 |
| inc0 | almost 400 on amazon | 14:58 |
| jrich523 | lol | 14:58 |
| sdake_ | on the plus side they wont ever break :) | 14:58 |
| jrich523 | im fairly certain my wife would kill me | 14:58 |
| sdake_ | and they sound fantastic | 14:59 |
| jrich523 | i bet | 14:59 |
| inc0 | I'm also looking at sennheiser 598 | 14:59 |
| inc0 | I have really good experience with sennheisers | 14:59 |
| sdake_ | sennheiser makes some good and not so good headphones | 14:59 |
| sdake_ | hard to pick em apart | 14:59 |
| jrich523 | i found the AT to be a bit more comfortable | 15:00 |
| jrich523 | however the longer cups are nice | 15:00 |
| jrich523 | they are pretty ridged too, which is why i like the foldable ones | 15:00 |
| jrich523 | i tend to have one cup off my ear, cant do that with the sennheiser | 15:00 |
| inc0 | ath looks great | 15:02 |
| inc0 | I'll need some mic to them and it seems to be a hit | 15:02 |
| jrich523 | link above | 15:03 |
| jrich523 | for v-moda boom mic | 15:03 |
| jrich523 | also you need an adapter | 15:03 |
| jrich523 | https://www.amazon.com/Josi-Minea-Converter-Headphone-Earphone/dp/B017VHX4Q6/ref=pd_sim_107_7?ie=UTF8&psc=1&refRID=AXPM4RW366ZMC87W2X5K | 15:04 |
| jrich523 | i'd probably get the 2 pack | 15:04 |
| jrich523 | i've had decent luck, but in reading they often break i guess | 15:04 |
| jrich523 | i have the AT and my wife has the Senn, we both have the adapter with v-moda mic | 15:04 |
| sdake_ | anyone know where the tox -e doc change is in the review queue | 15:05 |
| inc0 | so this vmoda mic+ath+adapter and you have mic on same cable as phones? | 15:06 |
| jrich523 | so, the openstack_release is commented out, which, im fairly certain i didnt change this since my last release, also it redeployed to the controller node fine(30), just not the new node (21) | 15:06 |
| jrich523 | yeah | 15:06 |
| inc0 | yeah I like this rig, I'll probably just copy it and get it on my own | 15:06 |
| jrich523 | the mic comes with a cable (on/off, vol control) and that plugs right in to the headphones with the adapter | 15:06 |
| inc0 | thanks a lot! | 15:06 |
| jrich523 | also that adapter i link fits nicely in there | 15:07 |
| jrich523 | you can probably find it cheaper straight from china, just make sure its that style | 15:07 |
| jrich523 | or, rather, that exact one, you can find tones that look like that, for like $1 | 15:07 |
| jrich523 | just you know.. shipping time from china and what not :) | 15:07 |
| inc0 | I want my headphones next week | 15:08 |
| inc0 | ;) | 15:08 |
| jrich523 | also that way you can just leave the mic attached to the computer and use the orginal cable to travel with | 15:08 |
| inc0 | already ordered, so that's done deal;) | 15:08 |
| jrich523 | lol nice | 15:08 |
| kbaegis | Easiest way to push without building? | 15:08 |
| inc0 | that's also good idea | 15:08 |
| kbaegis | I'm getting unexpected x509 errors | 15:08 |
| inc0 | kbaegis, docker push <<image>> | 15:08 |
| kbaegis | inc0 ty, I'll try that with a docker images <namespace>/* -q? | 15:09 |
| *** dwalsh has joined #openstack-kolla | 15:09 | |
| inc0 | I think you need to push the tags | 15:10 |
| jrich523 | sdake_: the openstack_release setting in globals is commented out | 15:11 |
| jrich523 | which is how i think i fixed that the first time | 15:11 |
| inc0 | jrich523, it's in ansible/global_variables/all.yml | 15:11 |
| inc0 | that's the default | 15:11 |
| kbaegis | They're already tagged from build | 15:11 |
| jrich523 | oh | 15:12 |
| kbaegis | I just need the image reference I think for docker push, so the above + itterator | 15:12 |
| inc0 | globals.yml is just override file | 15:12 |
| inc0 | so you need to: docker push 172.21.0.253:5000/kolla/ubuntu-source-base | 15:12 |
| inc0 | ok it's official, I don't know how to rally | 15:20 |
| inc0 | I'll wait for pbourke to dump the results | 15:21 |
| *** vbel has quit IRC | 15:21 | |
| jrich523 | the only rally i can do is rally race :) | 15:21 |
| *** vbel has joined #openstack-kolla | 15:21 | |
| inc0 | subaru for life | 15:22 |
| jrich523 | i use to own an STI back in the day, what a great car | 15:22 |
| inc0 | I had just regular impreza | 15:22 |
| inc0 | when I'll have overflow of money (unlikely) I'll get sti | 15:22 |
| jrich523 | id kinda like another... but... i'd probably get something different next time around | 15:23 |
| jrich523 | however i got a chance to rally the BRZ, much more fun than i expected | 15:23 |
| inc0 | brz in rally? I thought awd is a must | 15:23 |
| inc0 | I have mazda mx5 miata, so I may try it too;) | 15:24 |
| inc0 | similar class to brz, only super light and convertible;) | 15:24 |
| jrich523 | convertible is not well liked in rally :) | 15:24 |
| inc0 | yeah, dust will be your worst enemy | 15:25 |
| openstackgerrit | Steven Dake proposed openstack/kolla: Make tox -e docs work https://review.openstack.org/358247 | 15:25 |
| jrich523 | even for autocross/rallycross | 15:25 |
| jrich523 | well mostly its a roll hazard | 15:25 |
| inc0 | hehe that too | 15:25 |
| jrich523 | you can suck all the dust you want, they dont care | 15:25 |
| inc0 | anyway, I wouldn't risk miata out there | 15:25 |
| inc0 | my little girl would die | 15:26 |
| inc0 | jrich523, you live in US? | 15:26 |
| jrich523 | yeah, Seattle Washington | 15:26 |
| inc0 | because rally doesn't seem too popular here | 15:26 |
| jrich523 | its not really | 15:26 |
| inc0 | only this nascar crap where people go left... | 15:26 |
| jrich523 | which, its getting to be a bit more, which is why i stopped lol | 15:26 |
| jrich523 | lol yeah pretty stupid | 15:26 |
| sdake_ | inc0 put a +2 on these: https://review.openstack.org/329454 https://review.openstack.org/358157 | 15:26 |
| jrich523 | basically people just wait for a crash | 15:26 |
| sdake_ | thats the whole point of rally | 15:27 |
| jrich523 | it use to be that i could go do a rally cross for a day and it wasnt too much of a problem, then it got to be where we had so many people some days i'd get 2 runs and that was it, so i said screw it | 15:27 |
| inc0 | hell yeah, we have xenial | 15:27 |
| jrich523 | also, i had already beat the hell out of my sti lol | 15:27 |
| sdake_ | and more importantly a working ubuntu gate | 15:28 |
| sdake_ | jrich523 i always wanted an sti | 15:28 |
| sdake_ | but wife made me get a panamera instead | 15:28 |
| inc0 | sdake_, nah, good rally doesn't mean a crash...but it is totally skill dependant | 15:28 |
| jrich523 | its been the most fun car i've ever driven | 15:28 |
| inc0 | jrich523, try miata | 15:29 |
| jrich523 | i own a camry hybrid and 4runner now lol | 15:29 |
| jrich523 | and... im married | 15:29 |
| inc0 | it's not fast car | 15:29 |
| inc0 | but damn it's fun | 15:29 |
| jrich523 | you know, i was looking at a small convertable for a 'fun' car | 15:29 |
| jrich523 | was looking at the Z4 | 15:29 |
| inc0 | miata all the way | 15:29 |
| jrich523 | i just cant get myself to try a miata | 15:29 |
| inc0 | well, Z4 is cool, I test driven it | 15:29 |
| jrich523 | it just feels.... wrong | 15:29 |
| sdake_ | porsche gt2 is where i'm headed after kids are out of the house | 15:29 |
| inc0 | nah | 15:29 |
| jrich523 | lol im a dink | 15:29 |
| inc0 | it's a love of my life, right after my wife and cat | 15:30 |
| inc0 | I couldn't fit to 2015 | 15:30 |
| inc0 | but try 2016 model | 15:30 |
| jrich523 | maybe i'll give it a go | 15:30 |
| inc0 | they changed chasis | 15:30 |
| jrich523 | however i think that fun car was put to the side and now we're looking at a plot of land | 15:30 |
| jrich523 | mostly it was the image i didnt really like | 15:30 |
| inc0 | and all the issues I had with 2015 disapeared | 15:30 |
| jrich523 | i've mostly seen a lot of older women driving them | 15:30 |
| jrich523 | be back in a couple of min | 15:31 |
| sdake_ | inc0 - would like to see those reviews go in soon as they are past the gate ;) | 15:31 |
| inc0 | jrich523, https://www.youtube.com/watch?v=xrr_OQiBevU look how happy they look | 15:31 |
| inc0 | sdake_, waiting for gate and I'll ack | 15:31 |
| inc0 | looks good so fa | 15:31 |
| inc0 | r | 15:31 |
| sdake_ | soundsgood | 15:31 |
| inc0 | jrich523, also if you really are up for it, you can put custom turbo in it;) | 15:33 |
| openstackgerrit | Steven Dake proposed openstack/kolla: Make tox -e docs work https://review.openstack.org/358247 | 15:33 |
| inc0 | probably still end up cheaper than z4 | 15:33 |
| *** dwalsh has quit IRC | 15:34 | |
| sdake_ | inc0 have a second look at https://review.openstack.org/#/c/357520/2 | 15:39 |
| patchbot | sdake_: patch 357520 - kolla - Fix the location of Kolla configurations in the ev... | 15:39 |
| Pavo | morning everyone | 15:39 |
| sdake_ | and recommend reading the whole docs on install - they HAVE CHANGED | 15:39 |
| sdake_ | hey pavo | 15:39 |
| sdake_ | we recommend pip install kolla for operators, git clone for devs | 15:39 |
| inc0 | sdake_, I know, but we don't have etc_examples dir now | 15:40 |
| sdake_ | inc0 if you pip install that directory is where the etc files rae stored | 15:40 |
| sdake_ | the exmples that is | 15:40 |
| sdake_ | read irc log in commit message formore detail | 15:41 |
| inc0 | ok makes sense | 15:41 |
| inc0 | also +a xenials | 15:42 |
| inc0 | great job | 15:42 |
| inc0 | Jeffrey4l_, woot ^ | 15:42 |
| sdake_ | ya Jeffrey4l_ really got to the root cause of the ubuntu gate being broken | 15:42 |
| sdake_ | after 2+ months... | 15:42 |
| sdake_ | needle in a haystack | 15:43 |
| inc0 | you know what was that? | 15:43 |
| sdake_ | makes perfect sense now | 15:43 |
| sdake_ | yup | 15:43 |
| sdake_ | we were using wheels | 15:43 |
| sdake_ | wheel was compiled against so .18 | 15:43 |
| inc0 | do share plz | 15:43 |
| sdake_ | the python code was expecting .20 | 15:43 |
| sdake_ | so.18 | 15:43 |
| sdake_ | so.20 | 15:43 |
| sdake_ | incompatible - hence lbirary fails to load in gate | 15:44 |
| sdake_ | because of the wheel cache | 15:44 |
| sdake_ | we sstill use the pypi cche in teh gate | 15:44 |
| sdake_ | but the wheel cache is turned off now | 15:44 |
| inc0 | ok that makes sense indeed | 15:45 |
| kbaegis | Pavo morning | 15:45 |
| sdake_ | not quite sure how jeffrey4l figured that one out | 15:45 |
| sdake_ | but ya ;) | 15:45 |
| openstackgerrit | Merged openstack/kolla: Fix the location of Kolla configurations in the evaluation workflow https://review.openstack.org/357520 | 15:47 |
| openstackgerrit | Merged openstack/kolla: Fix Ubuntu source gate deployment https://review.openstack.org/358157 | 15:48 |
| *** v1k0d3n has joined #openstack-kolla | 15:49 | |
| *** matrohon has joined #openstack-kolla | 15:49 | |
| Pavo | morning kbaegis | 15:50 |
| jrich523 | inc0: i found a used Z4 for like 15k | 15:51 |
| inc0 | jrich523, but keep in mind that BMW after warranty ends can be much more expensive | 15:51 |
| inc0 | thats why they lose so much value with age | 15:51 |
| jrich523 | yeah, true | 15:51 |
| jrich523 | but, i think im going to go with a plot of land instead | 15:52 |
| jrich523 | want to get like 20+ acres in the mtns | 15:52 |
| inc0 | I'd go with plot too | 15:52 |
| jrich523 | land is getting crazy around here | 15:52 |
| jrich523 | so figured it would be worth getting it and holding on to it | 15:52 |
| jrich523 | use it for camping and what not for now and later build a retirement home on it | 15:53 |
| jrich523 | ok, so i think last time zhugaoxiao fixed my tag 3.0 issue, however, he did that before i saw what he was doing via webex | 15:54 |
| jrich523 | which means, i dont have a clue whats going on here | 15:54 |
| openstackgerrit | Merged openstack/kolla: Pin to Ubuntu 16.04 https://review.openstack.org/329454 | 15:57 |
| sdake_ | jrich523 - there are new documentation changes | 16:00 |
| sdake_ | jrich523 i'd recommend reading them | 16:00 |
| jrich523 | looking now | 16:01 |
| sdake_ | in the qucksstart.rst | 16:01 |
| sdake_ | look for "Attention:" | 16:01 |
| *** dave-mccowan has quit IRC | 16:02 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Remove log propagate configuration https://review.openstack.org/358251 | 16:03 |
| jrich523 | ok | 16:03 |
| jrich523 | i think i need to blow away what i have and start over | 16:04 |
| jrich523 | also i think the 'quick start' should probably assume a stable branch | 16:05 |
| jrich523 | and there should be a 'dev setup guide' | 16:06 |
| openstackgerrit | Steven Dake proposed openstack/kolla: Fix bandit gate jobs https://review.openstack.org/353942 | 16:07 |
| sdake_ | jrich523 - emergency problems first | 16:08 |
| sdake_ | jrich523 - making it perfect -> later | 16:08 |
| sdake_ | jrich523 our docs are undergoing a restructure we agreed upon at the midcycle - it has yet to complete or really be a priority | 16:13 |
| sdake_ | even though it seems to be everyones #1 pain point | 16:13 |
| jrich523 | i would say most folks want to learn openstack | 16:14 |
| jrich523 | and they come here because, according to #openstack this is the easiest to get going with | 16:14 |
| jrich523 | which, is why im here | 16:14 |
| *** patchbot has quit IRC | 16:15 | |
| Pavo | I would have to agree with jrich523 stable docs should already be done to the point of nothing should be changed, and dev docs should be a continuous document | 16:15 |
| *** daneyon has joined #openstack-kolla | 16:16 | |
| sdake_ | yes we know | 16:16 |
| sdake_ | feel free to write em :) | 16:16 |
| Pavo | I wouldn't say that now jrich523 I think packstack is easier than kolla for deployment but I also just started learning kolla, and really like the way it works, just have to understand it all | 16:16 |
| sdake_ | pavo kolla is naturallly a little more complex because it does multlinode | 16:17 |
| sdake_ | and doesn't spew dependencies and files all over your disks | 16:17 |
| jrich523 | which i like :) | 16:17 |
| sdake_ | (i.e. the container part) | 16:17 |
| jrich523 | yeah i really like that its docker/ansible | 16:17 |
| Pavo | I use packstack with multinode | 16:17 |
| Pavo | and agree jrich523 | 16:18 |
| sdake_ | did they add multinode to it? | 16:18 |
| Pavo | I have been using packstack for 3 nodes for about a year now | 16:18 |
| sdake_ | does packstack launch from one node to multi ? | 16:18 |
| sdake_ | or you pacstack each one | 16:18 |
| Pavo | yeah all dependencies have to be on all nodes thats true | 16:19 |
| Pavo | but I can run the packstack cmd with an aswerfile from a deployer node | 16:19 |
| sdake_ | does packstack have single source of truth - that is you configure it in one place and it deploys to multiple nodes? | 16:19 |
| Pavo | it can if you have all the dependencies on all nodes | 16:20 |
| *** daneyon has quit IRC | 16:20 | |
| sdake_ | k | 16:21 |
| Pavo | don't get me wrong here, I'm not bashing on kolla or anything, thats why I am here. I love the way kolla works, just have to get to know what all its capable of | 16:21 |
| Pavo | and so far, its amazing | 16:21 |
| Jeffrey4l_ | it is time to enable Ubuntu binary deploy on master now. https://review.openstack.org/358253 | 16:21 |
| Pavo | man I need to get this on some baremetal this vmware shit is killing me | 16:25 |
| jrich523 | lol | 16:25 |
| sdake_ | lol | 16:25 |
| sdake_ | inc0 Jeffrey4l_ can I get acks on this doc fix | 16:26 |
| Pavo | yeah gonna stop and just wait until I get to work monday | 16:26 |
| sdake_ | https://review.openstack.org/358247 | 16:26 |
| Jeffrey4l_ | sdake_, Patch in Merge Conflict, need a rebase now. | 16:26 |
| jrich523 | i would suggest, in the Attention section, to suggest a branch | 16:27 |
| sdake_ | right | 16:27 |
| inc0 | sdake_, you have merge conflict...on a doc patch | 16:29 |
| openstackgerrit | Steven Dake proposed openstack/kolla: Make tox -e docs work https://review.openstack.org/358247 | 16:31 |
| sdake_ | wierd | 16:31 |
| jrich523 | image-building.rst title should have customization, not customisation | 16:31 |
| sdake_ | jrich523 just fixing tox -e fails | 16:31 |
| sdake_ | not reformatting style | 16:32 |
| sdake_ | jrich523 blame emea for that oen :) | 16:32 |
| jrich523 | lol | 16:32 |
| jrich523 | oddly, when i looked it up, i didnt find that the spelling was wrong... but... every page just spelled it with a z... kinda weird | 16:33 |
| sdake_ | s commonly replaces z in europe | 16:34 |
| jrich523 | yeah, figured that was the case | 16:34 |
| sdake_ | i didn't want to hold up pbourke's review with a nitpick like htat ;) | 16:35 |
| jrich523 | fair enough! | 16:35 |
| sdake_ | we sort of agreed we would only nitpick the quickstart guide from here on out | 16:35 |
| sdake_ | to let the other docs develop faster | 16:35 |
| jrich523 | i assume the most hit page is the quickstart | 16:35 |
| sdake_ | no idea | 16:36 |
| sdake_ | i dont have acceess to stats | 16:36 |
| sdake_ | https://image-store.slidesharecdn.com/cf78ff3e-8673-4e67-bd11-64e80009695f-original.jpeg | 16:46 |
| jrich523 | do you have that printed out hanging in your office? :) | 16:47 |
| *** hieulq_ has quit IRC | 16:53 | |
| jrich523 | lol im trying to configure bind, and i just ran an nslookup on it, and it cause nslookup to crash | 16:55 |
| sdake_ | My favorite is when some 22-year-old fresh out of college in some established industry introduces himself as an SME. I want to reply, "Honey, the only thing you're an SME in is bonging beers." | 16:58 |
| sdake_ | https://www.linkedin.com/pulse/which-business-bullst-phrases-annoy-you-most-steve-blakeman?trk=hp-feed-article-title-comment | 16:58 |
| jrich523 | lol | 16:58 |
| sdake_ | there was this cat at cisco | 16:59 |
| jrich523 | the only one that annoys me is the think outside the box | 16:59 |
| sdake_ | he sent out a daily spam of his status report to over 1k people | 16:59 |
| sdake_ | was, as in past tense ;) | 17:00 |
| jrich523 | meow that is excessive | 17:00 |
| *** bootsha has joined #openstack-kolla | 17:02 | |
| *** bootsha has quit IRC | 17:05 | |
| sdake_ | imo status reports should only be given to those that ask for them | 17:06 |
| sdake_ | inc0 speaking of status reports where a re you with the osic gear | 17:08 |
| sdake_ | i want to try my hand at pandocing the json we have | 17:08 |
| inc0 | so I failed at generation of these;) | 17:18 |
| sdake_ | failed at pandoc | 17:22 |
| sdake_ | or failed at results generation on ceph? | 17:22 |
| openstackgerrit | Merged openstack/kolla: Make tox -e docs work https://review.openstack.org/358247 | 17:22 |
| inc0 | yay lvl70 on monk sdake_ | 17:23 |
| sdake_ | i see - hardly working then :) | 17:23 |
| inc0 | grinding like crazy | 17:24 |
| sdake_ | you can grind after the 31st ;) | 17:24 |
| Pavo | lol playing that right now | 17:24 |
| sdake_ | my toon is 930 paragon | 17:25 |
| inc0 | Pavo, diablo 3? | 17:25 |
| sdake_ | barb - nearly bis gear | 17:25 |
| sdake_ | i was top 100 in season 6 -) | 17:26 |
| Daviey_ | inc0: Are you planning to use OSIC tomorrow? | 17:26 |
| *** Daviey_ is now known as Daviey | 17:26 | |
| sdake_ | Daviey i dont hink he is planning to use it today | 17:27 |
| sdake_ | since he is grinding diablo 3 :) | 17:27 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Use dumb-init to manage the pid 1 process https://review.openstack.org/357015 | 17:27 |
| inc0 | you misjudge me sdake_ , I already did reporting | 17:27 |
| inc0 | just need to figure out rally | 17:27 |
| inc0 | Daviey, do you have plans to use it tomorrow? | 17:29 |
| *** kbaegis has quit IRC | 17:29 | |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: DO_NOT_MERGE: Use better URL for China https://review.openstack.org/329060 | 17:30 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Create the nova instances directory when need during starting https://review.openstack.org/357898 | 17:30 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Implement mongodb replication set cluster https://review.openstack.org/345751 | 17:30 |
| openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Sync the local_settings.j2 with horizon https://review.openstack.org/356471 | 17:30 |
| Daviey | inc0: I'd like to quickly do a deployment of storage and compute co-located on the full cluster to see how it performs | 17:32 |
| inc0 | Daviey, +1 to that, I'd love to see results myself | 17:33 |
| inc0 | ceph I presume? | 17:33 |
| Daviey | inc0: yes | 17:33 |
| *** williamcaban has joined #openstack-kolla | 17:35 | |
| *** kbaegis has joined #openstack-kolla | 17:36 | |
| *** sdake_ has quit IRC | 17:40 | |
| *** kbaegis has quit IRC | 17:41 | |
| *** kbaegis has joined #openstack-kolla | 17:42 | |
| openstackgerrit | Dave Walker proposed openstack/kolla: Fix bandit gate jobs https://review.openstack.org/353942 | 17:42 |
| Daviey | ^ fixed pep8 issues | 17:43 |
| inc0 | rerunning rally tests | 17:45 |
| *** kbaegis has quit IRC | 17:47 | |
| *** kbaegis has joined #openstack-kolla | 18:01 | |
| *** daneyon has joined #openstack-kolla | 18:16 | |
| *** daneyon has quit IRC | 18:20 | |
| kbaegis | Is designate ready under kolla? | 18:52 |
| *** kbaegis has quit IRC | 18:58 | |
| *** kbaegis has joined #openstack-kolla | 18:58 | |
| *** JoseMello has quit IRC | 19:00 | |
| inc0 | that is a good question | 19:05 |
| inc0 | we have containers, but not role | 19:06 |
| *** papacz1 has quit IRC | 19:08 | |
| *** v1k0d3n has quit IRC | 19:09 | |
| *** v1k0d3n has joined #openstack-kolla | 19:09 | |
| *** v1k0d3n has quit IRC | 19:13 | |
| *** v1k0d3n has joined #openstack-kolla | 19:13 | |
| *** salv-orl_ has joined #openstack-kolla | 19:15 | |
| *** salv-orlando has quit IRC | 19:19 | |
| *** v1k0d3n has quit IRC | 19:21 | |
| *** HyperJohnGraham has quit IRC | 20:01 | |
| *** daneyon has joined #openstack-kolla | 20:04 | |
| *** daneyon has quit IRC | 20:09 | |
| sbezverk_ | anybody hit this issue on the latest master? https://bugs.launchpad.net/neutron/+bug/1615283 | 20:17 |
| openstack | Launchpad bug 1615283 in neutron "neutron-openvswitch-agent backtrace at ovs_bridge.py" [Undecided,New] | 20:17 |
| *** v1k0d3n has joined #openstack-kolla | 20:23 | |
| *** Jeffrey4l_ has quit IRC | 20:35 | |
| *** Jeffrey4l_ has joined #openstack-kolla | 20:36 | |
| *** v1k0d3n has quit IRC | 20:40 | |
| *** v1k0d3n has joined #openstack-kolla | 20:40 | |
| *** v1k0d3n has quit IRC | 20:44 | |
| *** v1k0d3n has joined #openstack-kolla | 20:44 | |
| *** v1k0d3n has quit IRC | 20:51 | |
| *** papacz has joined #openstack-kolla | 20:57 | |
| *** v1k0d3n has joined #openstack-kolla | 21:11 | |
| *** pbourke has joined #openstack-kolla | 21:11 | |
| *** mark-casey has joined #openstack-kolla | 21:12 | |
| *** pbourke has quit IRC | 21:20 | |
| *** matrohon has quit IRC | 21:30 | |
| *** v1k0d3n has quit IRC | 21:31 | |
| *** v1k0d3n has joined #openstack-kolla | 21:31 | |
| *** v1k0d3n has quit IRC | 21:34 | |
| *** inc0 has quit IRC | 21:35 | |
| *** v1k0d3n has joined #openstack-kolla | 21:35 | |
| *** v1k0d3n has quit IRC | 21:42 | |
| *** dmueller-evoila has quit IRC | 21:43 | |
| *** sdake has joined #openstack-kolla | 21:47 | |
| sdake | powernap ftw | 21:47 |
| *** daneyon has joined #openstack-kolla | 21:53 | |
| *** daneyon has quit IRC | 21:58 | |
| *** jrich523_ has joined #openstack-kolla | 22:01 | |
| *** mark-casey has quit IRC | 22:03 | |
| *** ajpower has quit IRC | 22:03 | |
| *** manjeets_ has quit IRC | 22:03 | |
| *** mag009 has quit IRC | 22:03 | |
| *** williamcaban has quit IRC | 22:03 | |
| *** ntpttr___ has quit IRC | 22:03 | |
| *** jrich523 has quit IRC | 22:03 | |
| *** akscram has quit IRC | 22:03 | |
| *** zeus has quit IRC | 22:03 | |
| *** cmceniry has quit IRC | 22:03 | |
| *** ntpttr___ has joined #openstack-kolla | 22:03 | |
| *** williamcaban has joined #openstack-kolla | 22:04 | |
| *** v1k0d3n has joined #openstack-kolla | 22:04 | |
| *** mag009_ has joined #openstack-kolla | 22:05 | |
| *** cmceniry has joined #openstack-kolla | 22:05 | |
| *** zigo has quit IRC | 22:08 | |
| *** zeus has joined #openstack-kolla | 22:09 | |
| *** zeus is now known as Guest37793 | 22:09 | |
| *** ajpower has joined #openstack-kolla | 22:10 | |
| *** zigo has joined #openstack-kolla | 22:10 | |
| *** jeh has quit IRC | 22:14 | |
| *** akscram has joined #openstack-kolla | 22:16 | |
| *** jeh has joined #openstack-kolla | 22:17 | |
| *** jeh has quit IRC | 22:17 | |
| *** jeh has joined #openstack-kolla | 22:17 | |
| *** v1k0d3n has quit IRC | 22:24 | |
| *** v1k0d3n has joined #openstack-kolla | 22:25 | |
| *** v1k0d3n has quit IRC | 22:28 | |
| *** v1k0d3n has joined #openstack-kolla | 22:28 | |
| *** v1k0d3n has quit IRC | 22:36 | |
| wznoinsk | I started to think you sleep never sdake | 23:21 |
| sdake | 6 hrs a day atleast | 23:21 |
| wznoinsk | similar here, I would be handling it better if wouldn't have to mind kids that much (few more years) | 23:22 |
| wznoinsk | out of curiosity what would need to happen to have baremetal out of techpreview in kolla? | 23:23 |
| wznoinsk | I do get sunday aftermidnight (UTC+1) is not the best time to ask these ;-), we can leave till workday | 23:24 |
| sdake | wznoinsk i work sometimes on the weekend | 23:26 |
| sdake | what i want is to be able to specify a flavor | 23:26 |
| sdake | one type of flavor = bare metal | 23:26 |
| sdake | other type = virtual machines | 23:26 |
| sdake | and schedule them from one nova stack | 23:27 |
| sdake | another option is to have two nova stacks - one for bare metal one for virtual machines | 23:27 |
| sdake | with two differneent endpoints | 23:27 |
| sdake | but this second option is not as ideal for obvious reasons | 23:27 |
| wznoinsk | is the hypervisor's hostname an issue when having two nova computes (vm and ironic) on the same machine a problem ? | 23:28 |
| sdake | no - our current implementation is either all baremetal or all virtual machine | 23:28 |
| sdake | it doesn't let a mix occur | 23:28 |
| sdake | jpeeler worked on it for a long time and chouldn't get it to work | 23:28 |
| sdake | he said he didn't think it was possible to do | 23:29 |
| sdake | which makes me question the value of ironic if that were the case | 23:29 |
| sdake | so it makes me think that is not the case, he just wasn't able to sort it out | 23:29 |
| * wznoinsk thinks straight, it's Guiness that writes these sophistications | 23:29 | |
| sdake | if you look at our ironic implementation now, it hsa conditionals all over nova.conf.j2 | 23:30 |
| wznoinsk | sdake: I get the point, I'm the 3rd in a row to work on the ironic with kolla, 2 previous ones failed... not because of lack of skills but time | 23:30 |
| sdake | i dont know if what i want is possible | 23:31 |
| sdake | but it makes the whole thing user friendly | 23:31 |
| sdake | vs unuser friendly | 23:31 |
| wznoinsk | did jpeeler give any clues on what was the 'nein' point? | 23:31 |
| sdake | no he didn't | 23:31 |
| sdake | he is on irc during the weekdays typically | 23:32 |
| sdake | he has since moved on to ot her interests | 23:32 |
| sdake | but im sure he would help with what he knows | 23:32 |
| wznoinsk | sdake: me kind of person to do on its own, will check during the week with him tho | 23:33 |
| sdake | whatever works | 23:33 |
| sdake | do you know enough about ironic to know if what i want is a possibility? | 23:34 |
| wznoinsk | no, working with ironic only last 3-4 weeks to get our kolla deployed development cloud running, one nova stack seems user friendly alright, was thinking containers should be able to provide enough of separation so the nova_compute_ironic wouldn't clash with the VM nvoa_compute one | 23:35 |
| *** bmace has quit IRC | 23:48 | |
| *** bmace has joined #openstack-kolla | 23:49 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
