| *** grumpycat has quit IRC | 00:11 | |
| *** grumpycat has joined #openstack-ansible | 00:11 | |
| *** grumpycat has quit IRC | 00:30 | |
| *** grumpycat has joined #openstack-ansible | 00:31 | |
| *** arbrandes has quit IRC | 00:35 | |
| *** shoutm_ has joined #openstack-ansible | 00:43 | |
| *** shoutm has quit IRC | 00:45 | |
| *** grumpycat has quit IRC | 00:46 | |
| *** grumpycat has joined #openstack-ansible | 00:48 | |
| *** sdake_ has quit IRC | 00:52 | |
| *** arbrandes has joined #openstack-ansible | 00:53 | |
| *** shoutm_ has quit IRC | 01:03 | |
| *** shoutm has joined #openstack-ansible | 01:08 | |
| *** arbrandes has quit IRC | 01:35 | |
| *** arbrandes has joined #openstack-ansible | 01:55 | |
| *** cfarquhar has quit IRC | 01:57 | |
| *** woodard has joined #openstack-ansible | 01:57 | |
| *** shoutm_ has joined #openstack-ansible | 02:01 | |
| *** shoutm has quit IRC | 02:03 | |
| *** sdake has joined #openstack-ansible | 03:22 | |
| *** tlian has quit IRC | 03:38 | |
| openstackgerrit | Steve Lewis proposed stackforge/os-ansible-deployment: Add support for swift proxy config: sorting_method https://review.openstack.org/208817 | 03:43 |
|---|---|---|
| *** shoutm_ has quit IRC | 03:48 | |
| *** shoutm has joined #openstack-ansible | 03:59 | |
| *** sdake_ has joined #openstack-ansible | 04:06 | |
| *** sdake has quit IRC | 04:10 | |
| *** sdake_ is now known as sdake | 04:18 | |
| coolj | palendae: https://bugs.launchpad.net/openstack-ansible/+bug/1488315 | 04:30 |
| openstack | Launchpad bug 1488315 in openstack-ansible "The python-requests package is pulled in by apt via dependency" [Undecided,New] | 04:30 |
| *** fawadkhaliq has joined #openstack-ansible | 04:57 | |
| *** javeriak has joined #openstack-ansible | 05:13 | |
| *** shausy has joined #openstack-ansible | 05:25 | |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Container create/system tuning https://review.openstack.org/215905 | 05:45 |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Implement /usr/bin/env as the shebang in all bash scripts https://review.openstack.org/215906 | 05:45 |
| *** javeriak has quit IRC | 05:49 | |
| *** javeriak has joined #openstack-ansible | 05:58 | |
| *** javeriak has quit IRC | 06:06 | |
| *** serverascode has quit IRC | 06:12 | |
| *** rward has quit IRC | 06:12 | |
| *** rward has joined #openstack-ansible | 06:15 | |
| *** serverascode has joined #openstack-ansible | 06:16 | |
| *** javeriak has joined #openstack-ansible | 06:31 | |
| *** javeriak_ has joined #openstack-ansible | 06:34 | |
| *** fawadk has joined #openstack-ansible | 06:34 | |
| *** fawadkhaliq has quit IRC | 06:35 | |
| *** javeriak has quit IRC | 06:35 | |
| *** shoutm has quit IRC | 06:49 | |
| *** shoutm has joined #openstack-ansible | 06:51 | |
| *** javeriak has joined #openstack-ansible | 06:53 | |
| *** javeriak_ has quit IRC | 06:53 | |
| *** shausy has quit IRC | 07:02 | |
| *** shausy has joined #openstack-ansible | 07:03 | |
| *** fawadk has quit IRC | 07:43 | |
| *** gparaskevas has joined #openstack-ansible | 07:49 | |
| openstackgerrit | Merged stackforge/os-ansible-deployment: Use dict args for ceph_config slurp https://review.openstack.org/215551 | 07:53 |
| *** javeriak has quit IRC | 07:54 | |
| *** shoutm has quit IRC | 08:03 | |
| *** shoutm has joined #openstack-ansible | 08:05 | |
| cloudnull | morning | 08:15 |
| gparaskevas | morning | 08:22 |
| *** javeriak has joined #openstack-ansible | 08:32 | |
| evrardjp | morning everyoen | 08:38 |
| evrardjp | everyone* | 08:38 |
| openstackgerrit | Matt Thompson proposed stackforge/os-ansible-deployment: Use dict args for ceph_config slurp https://review.openstack.org/216608 | 08:43 |
| odyssey4me | hughsaunders care to inspect the backport of your patch? https://review.openstack.org/215903 | 09:00 |
| hughsaunders | mattt: would you mind reviewing https://review.openstack.org/#/c/215903/ | 09:02 |
| hughsaunders | its my patch so I probably shouldn't review it | 09:02 |
| mattt | sure i can have a look | 09:05 |
| *** fawadkhaliq has joined #openstack-ansible | 09:09 | |
| gparaskevas | odyssey4me: whats going on with my patch? :P jenkins? http://cdn.meme.am/instances2/500x/1543538.jpg | 09:09 |
| odyssey4me | gparaskevas hpcloud is too slow, so the patch runs the gauntlet hoping to either end up on rax or on a decent hpcloud host | 09:10 |
| gparaskevas | odyssey4me: ok ! thanks | 09:11 |
| prometheanfire | don't abandon my patches kthnx | 09:13 |
| prometheanfire | also, it seemed to be everyone elses opinion that the path needs to be direct | 09:14 |
| prometheanfire | anyway, sleep | 09:14 |
| openstackgerrit | Jean-Philippe Evrard proposed stackforge/os-ansible-deployment: Fixing haproxy-playbook fails when installing on multiple hosts https://review.openstack.org/215579 | 09:19 |
| odyssey4me | prometheanfire don't submit patches to multiple branches simultaneously - we've discussed this before | 09:23 |
| odyssey4me | evrardjp I think it's best to have the haproxy patch you did default to the same behaviour as now, leaving the implementation of the non-local bind to the deployer. | 09:24 |
| prometheanfire | I've been told that it's fine with the -1 workflow | 09:24 |
| prometheanfire | and told it's not fine | 09:24 |
| evrardjp | odyssey4me: ok | 09:25 |
| evrardjp | I'll change that | 09:25 |
| evrardjp | I was hesitating | 09:25 |
| odyssey4me | prometheanfire it's best to get the agreement on solving the solution before submitting the patches to the other branches - that's why the contrib guidelines specifically ask for a merge before backporting | 09:25 |
| evrardjp | I'll create a group_vars for that, this way we keep the haproxy role clean | 09:25 |
| prometheanfire | odyssey4me: there was agreement, by at least one core, on the issue | 09:26 |
| openstackgerrit | Jean-Philippe Evrard proposed stackforge/os-ansible-deployment: Adds the ability to provide user certificates to HAProxy https://review.openstack.org/215525 | 09:28 |
| prometheanfire | anyway, I really do need sleep, so nn | 09:28 |
| odyssey4me | prometheanfire for reference, please see the action items and discussion in the community meeting: http://eavesdrop.openstack.org/meetings/openstack_ansible_meeting/2015/openstack_ansible_meeting.2015-07-02-16.01.html | 09:32 |
| mattt | yay repo-clone-mirror.yml is gone | 09:32 |
| odyssey4me | mattt only from master, but yes | 09:33 |
| prometheanfire | odyssey4me: kk | 09:33 |
| openstackgerrit | Jean-Philippe Evrard proposed stackforge/os-ansible-deployment: [WIP] Fixing haproxy-playbook fails when installing on multiple hosts https://review.openstack.org/215579 | 09:36 |
| *** shoutm has quit IRC | 09:41 | |
| openstackgerrit | Matt Thompson proposed stackforge/os-ansible-deployment: Update ceilometer_git_repo to use git.openstack.org https://review.openstack.org/216644 | 09:51 |
| mattt | github issues :( | 09:51 |
| *** grumpycat has quit IRC | 09:56 | |
| *** grumpycat has joined #openstack-ansible | 09:56 | |
| *** grumpycat is now known as Guest32429 | 09:56 | |
| *** javeriak has quit IRC | 10:00 | |
| odyssey4me | gparaskevas yup, github appears to be inaccessible from hpcloud at this time | 10:09 |
| odyssey4me | not sure about rax right now | 10:09 |
| gparaskevas | odyssey4me: yeah i saw that :/ | 10:10 |
| gparaskevas | odyssey4me: thanks | 10:10 |
| *** javeriak has joined #openstack-ansible | 10:25 | |
| openstackgerrit | Jean-Philippe Evrard proposed stackforge/os-ansible-deployment: Fixing haproxy-playbook fails when installing on multiple hosts https://review.openstack.org/215579 | 10:25 |
| openstackgerrit | Jean-Philippe Evrard proposed stackforge/os-ansible-deployment: Fixing haproxy-playbook fails when installing on multiple hosts https://review.openstack.org/215579 | 10:36 |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Enable tempest testing of ceilometer https://review.openstack.org/209568 | 10:43 |
| *** britthouser has quit IRC | 10:45 | |
| *** britthouser has joined #openstack-ansible | 10:47 | |
| *** javeriak_ has joined #openstack-ansible | 11:07 | |
| *** gparaskevas has quit IRC | 11:09 | |
| *** javeriak has quit IRC | 11:10 | |
| *** javeriak has joined #openstack-ansible | 11:16 | |
| *** javeriak_ has quit IRC | 11:17 | |
| *** javeriak has quit IRC | 11:20 | |
| *** fawadkhaliq has quit IRC | 11:20 | |
| *** javeriak has joined #openstack-ansible | 11:30 | |
| *** javeriak has quit IRC | 11:38 | |
| *** javeriak has joined #openstack-ansible | 11:52 | |
| *** javeriak has quit IRC | 11:59 | |
| *** Guest32429 is now known as grumpycatt | 12:02 | |
| *** shoutm has joined #openstack-ansible | 12:02 | |
| *** grumpycatt is now known as grumpylama | 12:03 | |
| *** javeriak has joined #openstack-ansible | 12:04 | |
| *** woodard has quit IRC | 12:05 | |
| *** woodard has joined #openstack-ansible | 12:05 | |
| *** fawadkhaliq has joined #openstack-ansible | 12:10 | |
| *** fawadkhaliq has quit IRC | 12:14 | |
| *** fawadkhaliq has joined #openstack-ansible | 12:20 | |
| *** prad_ has joined #openstack-ansible | 12:22 | |
| *** prad_ is now known as pradk | 12:24 | |
| odyssey4me | evrardjp is the overview on https://launchpad.net/openstack-ansible better now? | 12:50 |
| *** k_stev has joined #openstack-ansible | 12:56 | |
| *** mhayden has joined #openstack-ansible | 12:59 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 13:00 | |
| *** tlian has joined #openstack-ansible | 13:02 | |
| *** rrrobbb has joined #openstack-ansible | 13:06 | |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Removed default lxc profile on container create https://review.openstack.org/216301 | 13:15 |
| *** scarlisle has joined #openstack-ansible | 13:30 | |
| svg | https://review.openstack.org/#/c/215897/ sounds like low hanging fruit | 13:31 |
| * svg is updating his kilo test branch | 13:32 | |
| *** bapalm has quit IRC | 13:44 | |
| *** mpmsimo has joined #openstack-ansible | 13:49 | |
| *** bapalm_ is now known as bapalm | 13:51 | |
| *** javeriak has quit IRC | 13:52 | |
| *** KLevenstein has joined #openstack-ansible | 13:52 | |
| *** daneyon_ has quit IRC | 13:58 | |
| *** daneyon has joined #openstack-ansible | 13:58 | |
| *** spotz_zzz is now known as spotz | 13:59 | |
| *** Mudpuppy has joined #openstack-ansible | 14:05 | |
| *** spotz is now known as spotz_zzz | 14:08 | |
| *** yaya has joined #openstack-ansible | 14:12 | |
| *** mpmsimo has quit IRC | 14:15 | |
| *** spotz_zzz is now known as spotz | 14:24 | |
| *** yaya has quit IRC | 14:38 | |
| evrardjp | odyssey4me: sorry I was busy ... yes the overview seems perfect now | 14:42 |
| evrardjp | I've done a branding role for us, should I share it? I don't think it's really relevant for OSAD | 14:43 |
| *** yaya has joined #openstack-ansible | 14:43 | |
| *** jwagner_away is now known as jwagner | 14:50 | |
| *** rrrobbb has quit IRC | 14:52 | |
| *** shoutm has quit IRC | 14:52 | |
| *** k_stev has quit IRC | 14:56 | |
| *** javeriak has joined #openstack-ansible | 14:59 | |
| *** javeriak has quit IRC | 15:01 | |
| *** javeriak has joined #openstack-ansible | 15:01 | |
| *** fawadkhaliq has quit IRC | 15:15 | |
| *** javeriak has quit IRC | 15:21 | |
| *** javeriak has joined #openstack-ansible | 15:24 | |
| *** alop has joined #openstack-ansible | 15:27 | |
| *** fawadkhaliq has joined #openstack-ansible | 15:27 | |
| *** shausy has quit IRC | 15:30 | |
| openstackgerrit | Merged stackforge/os-ansible-deployment: Update ceilometer_git_repo to use git.openstack.org https://review.openstack.org/216644 | 15:31 |
| palendae | evrardjp: We have one in our rpc-openstack repo | 15:32 |
| cloudnull | o/ | 15:33 |
| *** woodard has quit IRC | 15:41 | |
| *** woodard has joined #openstack-ansible | 15:41 | |
| *** woodard has quit IRC | 15:42 | |
| *** woodard has joined #openstack-ansible | 15:42 | |
| *** cfarquhar has joined #openstack-ansible | 15:57 | |
| *** cfarquhar has quit IRC | 15:57 | |
| *** cfarquhar has joined #openstack-ansible | 15:57 | |
| *** yaya has quit IRC | 15:57 | |
| Sam-I-Am | y hello bug triage | 16:00 |
| odyssey4me | :) | 16:02 |
| cloudnull | good day gents | 16:02 |
| * odyssey4me hands the baton to cloudnull | 16:02 | |
| prometheanfire | if you say so | 16:02 |
| cloudnull | so im going to try and remember how to do this | 16:03 |
| hughsaunders | welcome back peateaell | 16:03 |
| cloudnull | :) | 16:03 |
| cloudnull | first up https://bugs.launchpad.net/openstack-ansible/+bug/1488532 | 16:03 |
| openstack | Launchpad bug 1488532 in openstack-ansible "lvm.conf template overwrites local configs" [Undecided,New] | 16:03 |
| cloudnull | thanks hughsaunders :) | 16:03 |
| *** yaya has joined #openstack-ansible | 16:04 | |
| cloudnull | so this looks like something that we should tackle sooner than later for the ceph work going on | 16:04 |
| cloudnull | IE loosing the lvm config on a host causing ceph downtime would be a bad thing . imo | 16:04 |
| odyssey4me | yep | 16:05 |
| odyssey4me | that's pretty nasty | 16:05 |
| cloudnull | so confirmed , medium for now. | 16:05 |
| odyssey4me | do we need to overwrite the file every time, or should we perhaps do something like a line-in-file instead? | 16:05 |
| cloudnull | the lvm conf is a template as I recal | 16:05 |
| hughsaunders | https://github.com/stackforge/os-ansible-deployment/blob/master/playbooks/roles/os_cinder/templates/lvm.conf.j2 | 16:05 |
| *** javeriak has quit IRC | 16:05 | |
| cloudnull | so we likley need to add a conditional there. | 16:05 |
| odyssey4me | yep, but should it be | 16:06 |
| hughsaunders | its a bit too involved to be line-in-filed | 16:06 |
| odyssey4me | also, do other roles also drop a template and maybe conflict? | 16:06 |
| cloudnull | i think so , but perhaps someone working on the ceph bits should make that call | 16:06 |
| cloudnull | do we want to target that for 11.2.0/1 ? | 16:07 |
| odyssey4me | dunno - it depends if someone can pick it up in the next few days | 16:08 |
| odyssey4me | it'll have to be 11.2.1 at the earliest, unless it's critical | 16:08 |
| cloudnull | ok | 16:08 |
| odyssey4me | we have a ton of backports which aren't getting through yet | 16:08 |
| cloudnull | 11.2.1 it is | 16:08 |
| odyssey4me | and 11.2.0 is friday | 16:08 |
| odyssey4me | I would default any milestones from this session to 11.2.1 - if someone gets a patch in we can pull it back. | 16:09 |
| cloudnull | next https://bugs.launchpad.net/openstack-ansible/+bug/1488161 | 16:10 |
| openstack | Launchpad bug 1488161 in openstack-ansible "Ansible should log to file" [Undecided,New] | 16:10 |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Switch to using git.openstack.org for git sources https://review.openstack.org/216780 | 16:10 |
| Sam-I-Am | ^ lulz | 16:11 |
| odyssey4me | uh, well we only set it to log to a file in a gate check nowadays | 16:11 |
| cloudnull | i think is something that a deployer can do without our intervention. imo | 16:11 |
| odyssey4me | and that's a deployer's choice which is easily set if desired | 16:11 |
| Sam-I-Am | true | 16:11 |
| cloudnull | it can be set locally or globally. | 16:11 |
| Sam-I-Am | just seems all that stuff scrolls by so quickly, its difficult in all cases to debug issues | 16:11 |
| Sam-I-Am | so by default it might be sort of nice to put the output someplace | 16:12 |
| cloudnull | maybe we doc how to do it ? or doc to the upstream docs on how its done? | 16:12 |
| odyssey4me | cloudnull but this is a standard ansible ability | 16:12 |
| odyssey4me | we should not rewrite ansible docs | 16:12 |
| cloudnull | right. | 16:12 |
| odyssey4me | this is like having us document how to create a cinder volume | 16:13 |
| Sam-I-Am | in the upstream openstack docs, we do things like 'optional: enable verbose mode in config file' | 16:13 |
| Sam-I-Am | but its up to the deployer how to figure it out | 16:13 |
| Sam-I-Am | so we could do it as just a reminder 'hey this might make life easier' | 16:13 |
| cloudnull | we can add this link to our docs http://docs.ansible.com/ansible/intro_configuration.html#log-path | 16:13 |
| Sam-I-Am | problem is, a lot of our new users are probably not ansible people | 16:13 |
| Sam-I-Am | in fact, from sitting in #openstack, they're not | 16:14 |
| Sam-I-Am | they like what we do, but usually come from chef or something | 16:14 |
| odyssey4me | While I would like to make life easier for our users, I do feel that anyone who wants to work with openstack at all should be able to figure this option out. | 16:14 |
| odyssey4me | we can perhaps refer to the ansible docs in some sort of quickstart. | 16:15 |
| cloudnull | +1 for that | 16:15 |
| Sam-I-Am | what the bug really says is... does it make sense to log by default vs. not? | 16:15 |
| openstackgerrit | Merged stackforge/os-ansible-deployment: Enable HAProxy Stats Web UI https://review.openstack.org/215019 | 16:15 |
| odyssey4me | Sam-I-Am I am -1 on logging ansible by default | 16:15 |
| cloudnull | imho , not. | 16:15 |
| odyssey4me | we do it for the gate only | 16:16 |
| cloudnull | next https://bugs.launchpad.net/openstack-ansible/+bug/1488315 | 16:17 |
| openstack | Launchpad bug 1488315 in openstack-ansible "The python-requests package is pulled in by apt via dependency" [Undecided,New] | 16:17 |
| Sam-I-Am | Apsu: trace city on that 3.16 box | 16:18 |
| Sam-I-Am | every few seconds | 16:18 |
| odyssey4me | hmm, is this an issue - if the requests conflicts with the wheels then it'll just be removed and replaced | 16:19 |
| palendae | It's an apt pacakge | 16:20 |
| palendae | package* | 16:20 |
| palendae | Not a wheel | 16:20 |
| cloudnull | so with the request bit, the wheel, if newer will replace it. | 16:20 |
| palendae | The wheel is not newer in 10.x | 16:20 |
| palendae | The apt packages pull in like 2.6 something | 16:21 |
| *** spotz is now known as spotz_zzz | 16:21 | |
| *** spotz_zzz is now known as spotz | 16:21 | |
| cloudnull | in 11 i dont see the apt package present | 16:21 |
| cloudnull | https://github.com/stackforge/os-ansible-deployment/blob/kilo/playbooks/roles/lxc_hosts/defaults/main.yml | 16:21 |
| palendae | No, not specifically | 16:22 |
| odyssey4me | also, I've seen the wheel downgrade it (I think) | 16:22 |
| palendae | Pulled in by lxc-templates | 16:22 |
| palendae | It's not added directly, it's pulled in via several dependencies, and it's breaking installs | 16:23 |
| odyssey4me | it is? | 16:23 |
| palendae | Read the bug | 16:23 |
| palendae | lxc-templates -> cloud-image-utils -> euac2ools -> python-requests | 16:24 |
| odyssey4me | ok -so let me check my understanding | 16:24 |
| odyssey4me | this is pulled in on hosts which have lxc installed, right? | 16:24 |
| palendae | Yes | 16:25 |
| odyssey4me | this equates the the controllers only | 16:25 |
| palendae | Nope | 16:25 |
| palendae | Logger runs lxc too | 16:25 |
| odyssey4me | the nova and glance clients aren't run on those hosts except in situations where it's a lab or an operator runs them | 16:25 |
| odyssey4me | so this has minimal impact if I understand the issue correctly - it only affects operators | 16:26 |
| palendae | 'only affects operators' | 16:26 |
| palendae | Who are users of our software | 16:26 |
| palendae | I don't see that as being minimal | 16:26 |
| odyssey4me | so we specifically setup a utility container for administration | 16:27 |
| odyssey4me | this is a bug and we should figure out how to deal with it, but it doesn't break any services | 16:28 |
| cloudnull | cfarquhar: is jordan around ? | 16:29 |
| odyssey4me | max I would say is medium, as there is a known workaround | 16:30 |
| cloudnull | so confirmed and marked med | 16:31 |
| odyssey4me | I've added a comment with the workaround info. | 16:31 |
| cloudnull | next https://bugs.launchpad.net/openstack-ansible/+bug/1488578 | 16:32 |
| openstack | Launchpad bug 1488578 in openstack-ansible "Add defaults for horizon ssl certs into upgrade.sh" [Undecided,New] | 16:32 |
| *** richoid has quit IRC | 16:32 | |
| cloudnull | this looks like a potential upgrade issue. | 16:32 |
| cloudnull | this seems like a deployment decision . we can doc it as something that can be set however its not something that breaks anything. | 16:33 |
| odyssey4me | this is a deployer choice - all the keystone/horizon plays will do is regenerate fresh certs so I don't see this as an issue | 16:33 |
| odyssey4me | especially since they're self-signed | 16:34 |
| cloudnull | it creats new , more secure certs than the generated ones provided in juno. | 16:34 |
| cloudnull | bjorne around ? | 16:34 |
| palendae | The docs were incredibly unclear | 16:34 |
| palendae | Basically, if you don't want a regeneration of the certs, set the vars to the old values | 16:35 |
| cloudnull | yes. | 16:35 |
| palendae | rpc has docs, but they need to be reworded | 16:35 |
| cloudnull | so can we improve the docs? Sam-I-Am? | 16:35 |
| palendae | Whether or not the defaults are automatically remapped is a separate issue | 16:35 |
| cloudnull | kk. confirmed, low, assigned to docs. | 16:36 |
| *** richoid has joined #openstack-ansible | 16:37 | |
| cloudnull | last on the new front: https://bugs.launchpad.net/openstack-ansible/+bug/1488156 | 16:37 |
| openstack | Launchpad bug 1488156 in openstack-ansible " Keystone SSL cert/key distribution and configuration" [Undecided,New] | 16:37 |
| cloudnull | which is a docimpact issue. | 16:37 |
| odyssey4me | yup | 16:37 |
| odyssey4me | low & assign to docs | 16:38 |
| cloudnull | done | 16:38 |
| cloudnull | what else we got? | 16:38 |
| Sam-I-Am | yeah, the last one... think we can fix docs | 16:38 |
| Sam-I-Am | problem is we dont have enough people testing these things | 16:38 |
| cloudnull | anything that we want to raise while we're all here ? | 16:38 |
| Sam-I-Am | the roof? | 16:39 |
| openstackgerrit | Kevin Carter proposed stackforge/os-ansible-deployment: [WIP] Replaced the copy_update module https://review.openstack.org/216790 | 16:39 |
| *** Bjoern_ has joined #openstack-ansible | 16:39 | |
| cloudnull | ok well then we're done here. | 16:40 |
| cloudnull | thanks all | 16:40 |
| *** Bjoern_ is now known as BjoernT | 16:42 | |
| *** richoid has quit IRC | 16:45 | |
| *** richoid has joined #openstack-ansible | 16:51 | |
| *** javeriak has joined #openstack-ansible | 16:58 | |
| *** abitha has joined #openstack-ansible | 16:58 | |
| *** javeriak has quit IRC | 17:03 | |
| *** javeriak has joined #openstack-ansible | 17:05 | |
| *** k_stev has joined #openstack-ansible | 17:13 | |
| *** jwagner is now known as jwagner_lunch | 17:23 | |
| *** yaya has quit IRC | 17:24 | |
| *** k_stev has quit IRC | 17:54 | |
| *** gparaskevas has joined #openstack-ansible | 17:55 | |
| *** yaya has joined #openstack-ansible | 17:55 | |
| gparaskevas | odyssey4me: Merged! thanks Jesse, looking forward for my next contribution! | 17:55 |
| *** alop has quit IRC | 18:09 | |
| *** britthouser has quit IRC | 18:17 | |
| *** britthouser has joined #openstack-ansible | 18:17 | |
| *** sigmavirus24 has quit IRC | 18:17 | |
| *** _sigmavirus24 has joined #openstack-ansible | 18:19 | |
| *** _sigmavirus24 is now known as sigmavirus24 | 18:20 | |
| *** sigmavirus24 has joined #openstack-ansible | 18:20 | |
| *** yaya has quit IRC | 18:24 | |
| *** sdake_ has joined #openstack-ansible | 18:25 | |
| *** sdake has quit IRC | 18:29 | |
| *** gparaskevas has quit IRC | 18:35 | |
| *** jwagner_lunch is now known as jwagner | 18:40 | |
| *** yaya has joined #openstack-ansible | 18:45 | |
| *** daneyon_ has joined #openstack-ansible | 18:53 | |
| *** sdake_ is now known as sdake | 18:54 | |
| *** daneyon has quit IRC | 18:55 | |
| *** KLevenstein_ has joined #openstack-ansible | 19:00 | |
| *** KLevenstein has quit IRC | 19:00 | |
| *** KLevenstein_ is now known as KLevenstein | 19:00 | |
| *** daneyon_ has quit IRC | 19:12 | |
| *** abitha has quit IRC | 19:12 | |
| *** daneyon has joined #openstack-ansible | 19:12 | |
| *** daneyon has quit IRC | 19:12 | |
| *** sdake_ has joined #openstack-ansible | 19:15 | |
| *** sdake has quit IRC | 19:18 | |
| *** sdake_ is now known as sdake | 19:23 | |
| openstackgerrit | Major Hayden proposed stackforge/os-ansible-deployment: Add profiling for Ansible tasks https://review.openstack.org/216849 | 19:31 |
| palendae | mhayden: Do you have an issue for that? We'd like to have LP bugs for stuff if we can | 19:32 |
| mhayden | palendae: i don't -- i'll go make one | 19:32 |
| palendae | Sweet, thanks! | 19:32 |
| mhayden | palendae: feel free to let me know if i skip steps -- figured i'll skin my knees as i go here ;) | 19:33 |
| palendae | np | 19:33 |
| palendae | We don't have a great contributing guide yet, either | 19:33 |
| palendae | But the LP bugs help us better track things against milestones | 19:33 |
| * mhayden notices there is a contributing.rst now | 19:33 | |
| palendae | And you can then add Closes-Bug: #<whatever> to have gerrit link up to LP | 19:33 |
| * mhayden hides in the corner ;) | 19:33 | |
| palendae | Oh there is? | 19:33 |
| mhayden | https://github.com/stackforge/os-ansible-deployment/blob/master/CONTRIBUTING.rst | 19:34 |
| palendae | Too much time in upgrade land | 19:34 |
| palendae | Oh right | 19:34 |
| palendae | That | 19:34 |
| openstackgerrit | Major Hayden proposed stackforge/os-ansible-deployment: Add profiling for Ansible tasks https://review.openstack.org/216849 | 19:36 |
| palendae | mhayden: Thanks! | 19:37 |
| mhayden | palendae: whoa, i did a thing! | 19:37 |
| palendae | I think there's a bug where the gerrit bot doesn't do the right link up stuff for launchpad if closes-bug is added in patchsets after 1...but this at least gets a topic and a link back from the commit | 19:37 |
| mhayden | ah okay | 19:38 |
| mhayden | i'll make an effort to get a bug in first | 19:38 |
| *** fawadkhaliq has quit IRC | 19:50 | |
| *** cloudtrainme has joined #openstack-ansible | 19:56 | |
| sigmavirus24 | palendae: that's the truth | 20:00 |
| sigmavirus24 | It's only on the initial review creation event | 20:00 |
| * sigmavirus24 thinks mhayden should go fix that in openstack/jeepyb | 20:00 | |
| sigmavirus24 | or openstack-dev | 20:00 |
| sigmavirus24 | or what | 20:00 |
| sigmavirus24 | *whatever | 20:00 |
| palendae | Pretty sure that's an OSIC thing | 20:00 |
| palendae | :) | 20:00 |
| *** javeriak has quit IRC | 20:01 | |
| sigmavirus24 | pretty sure you're an osic thing | 20:02 |
| *** javeriak has joined #openstack-ansible | 20:05 | |
| mhayden | sigmavirus24: looking now | 20:12 |
| mhayden | sigmavirus24: this one? https://github.com/openstack-infra/jeepyb | 20:13 |
| palendae | Uh oh | 20:13 |
| palendae | mhayden's like sigmavirus24 | 20:13 |
| palendae | "This project has a bug." "k, fixing." | 20:14 |
| palendae | Looks like the right one to me | 20:14 |
| mhayden | haha, oh no | 20:14 |
| mhayden | what did i step in now? | 20:14 |
| palendae | Though I've not messed with many openstack-infra projects outside of project-config | 20:14 |
| mhayden | oh, so this is the code that makes the gerrit <-> LP bridge? | 20:14 |
| palendae | Yeah | 20:14 |
| *** KLevenstein has quit IRC | 20:15 | |
| mhayden | the alarm by the revolving door really weirds me out | 20:17 |
| mhayden | i figure i'll quit jumping by the end of the week | 20:17 |
| palendae | mhayden: Evidently wherever our team has sat for the last few months (both moves) has had a TON of people going in and out of alarmed doors | 20:17 |
| mhayden | it's alarming | 20:18 |
| sigmavirus24 | Sorry | 20:18 |
| sigmavirus24 | mhayden: yes | 20:18 |
| sigmavirus24 | mhayden: we'll get along well you and me | 20:18 |
| *** cloudtrainme has quit IRC | 20:19 | |
| *** KLevenstein has joined #openstack-ansible | 20:22 | |
| *** cloudtrainme has joined #openstack-ansible | 20:27 | |
| palendae | mhayden: Is your profileing code from a version of https://github.com/ansible/ansible/blob/devel/lib/ansible/plugins/callback/profile_tasks.py ? | 20:28 |
| mhayden | palendae: i got mine from what was in the fedora project's infrastructure | 20:28 |
| palendae | Ok, was it apache-licensed? | 20:29 |
| mhayden | i'd imagine i could rewrite it | 20:29 |
| mhayden | let me go check the license on it | 20:29 |
| palendae | I ask because we've gotten submissions that were straight up cp's of upstream ansible modules, which are GPL :( | 20:29 |
| mhayden | ugh, didn't realize it was GPLv3 on the ansible side | 20:29 |
| mhayden | that's sad | 20:29 |
| openstackgerrit | Erik Wilson proposed stackforge/os-ansible-deployment: Archive Keystone to Keystone Federation rst content https://review.openstack.org/211747 | 20:30 |
| palendae | Yeah, Ansible's plugin system leaves some things to be desired | 20:30 |
| mhayden | source: https://infrastructure.fedoraproject.org/cgit/ansible.git/tree/callback_plugins/profile_tasks.py | 20:30 |
| mhayden | no license on any of it -- i'll ask the fedora infra folks | 20:30 |
| palendae | k, thanks | 20:30 |
| cloudnull | ok im off to sleep now. have a good one gents | 20:34 |
| palendae | Haven't looked really deeply yet but the file I saw was on Ansible 2.0. 1.9 doesn't have a profile callback plugin | 20:34 |
| palendae | cloudnull: Night | 20:34 |
| mhayden | palendae: since it's unlicensed, the FPCA default license says it's MIT licensed -> https://fedoraproject.org/wiki/Legal:Fedora_Project_Contributor_Agreement | 20:36 |
| mhayden | search for "Current Default License" on that page | 20:36 |
| mhayden | palendae: will add that to ticket | 20:37 |
| palendae | IANAL but I think MIT and Apache2 (what we, and I believe the rest of OpenStack use) are compatible for shipping | 20:37 |
| palendae | I don't know where that stands with big tent | 20:37 |
| mhayden | looks like MIT is okay -> https://wiki.openstack.org/wiki/Requirements | 20:38 |
| mhayden | along with BSD and Apache2 | 20:38 |
| palendae | Ok | 20:38 |
| * mhayden learned another new thing today | 20:39 | |
| *** abitha has joined #openstack-ansible | 20:40 | |
| mhayden | palendae: should i change the comments at the top of the file to reflect the MIT license? | 20:40 |
| palendae | Probably. I'm paranoid about getting stuck with a relicensing problem | 20:41 |
| mhayden | will do | 20:41 |
| evrardjp | palendae: I think you're paranoid about that | 20:47 |
| evrardjp | because every contribution is given to the foundation | 20:47 |
| palendae | evrardjp: Sure, but I'd rather be safe than sorry :) | 20:47 |
| openstackgerrit | Major Hayden proposed stackforge/os-ansible-deployment: Add profiling for Ansible tasks https://review.openstack.org/216849 | 20:47 |
| palendae | With licenses, since software developers != lawyers | 20:47 |
| palendae | Even though they love to think they are | 20:47 |
| evrardjp | it's mentionned it's welcomed in the dev guide | 20:47 |
| evrardjp | IIRC | 20:47 |
| palendae | Sure | 20:47 |
| xar- | odyssey4me: ping | 20:49 |
| evrardjp | software devs like to think they are anything | 20:49 |
| evrardjp | even unicorns! | 20:49 |
| palendae | Correct | 20:49 |
| palendae | Software dev ========= subject matter expert on * | 20:50 |
| palendae | But again, I'd rather be careful in terms of licenses. That's just me | 20:50 |
| evrardjp | https://wiki.openstack.org/wiki/LegalIssuesFAQ | 20:51 |
| *** javeriak has quit IRC | 20:51 | |
| evrardjp | that's what I mentionned | 20:51 |
| evrardjp | mentionned, is it the correct term in english? | 20:51 |
| evrardjp | I meant referenced to | 20:51 |
| palendae | Yep | 20:51 |
| palendae | Ok, the incorporating MIT/BSD one is pretty much what I'd expect | 20:51 |
| palendae | mhayden: ^^^^ | 20:52 |
| palendae | So the new header looks right to me | 20:52 |
| evrardjp | I'm impressed to see the mhayden here, it's been ages that I first see his blog major.io | 20:53 |
| palendae | evrardjp: You'll be seeing him a lot more, I think | 20:54 |
| palendae | He's on our team at RAX now | 20:54 |
| evrardjp | cool :) | 20:55 |
| evrardjp | chances are low though | 20:55 |
| evrardjp | I'm not in the same timezone | 20:55 |
| evrardjp | I shouldn't be here right now | 20:55 |
| palendae | It's only 3pm there right now | 20:55 |
| palendae | Ah, yeah | 20:55 |
| palendae | Eastern Europe? | 20:55 |
| evrardjp | it's 11PM | 20:55 |
| evrardjp | almost | 20:55 |
| evrardjp | that could change in the future, nobody knows | 20:56 |
| palendae | xar-: Guessing odyssey4me is out, it's ~10pm where he's at | 20:56 |
| *** daneyon has joined #openstack-ansible | 20:56 | |
| xar- | oh ok, I attempted to use OSAD late last week, failed miserably, I had some concerns regarding the configure_diskspace() bash function performing a mkpart on secondary block device | 20:58 |
| xar- | also, it didn't observe that I had available physical extents in a pre-existing volume group available for allocation | 20:59 |
| palendae | That's largely for an all-in-one/gates/VMs for testing, I think. How were you running it? And were you on physical hardware? | 20:59 |
| xar- | yup AIO lab environment, dedicated gear | 20:59 |
| palendae | k | 21:00 |
| palendae | So a physical AIO? | 21:00 |
| xar- | my terminology is shaky, but I believe so, yes, it's Ubuntu 14.04.3 LTS on the bare metal | 21:00 |
| palendae | The AIO scripts make a lot of assumptions, mostly that they're running on VMs that resemble the HP/RAX nodes in openstack-infra | 21:00 |
| evrardjp | I'm off for today, I think | 21:00 |
| palendae | Yeah, I'd call that a physical AIO; I have no doubt those scripts could made more intelligent | 21:00 |
| xar- | well I created a github account, I'm going to take a crack at rewriting scripts-library.sh to make it do what I need; seperate issue, it failed hard on the lxc-create portion | 21:01 |
| xar- | outside of AIO, is there an alternative to getting up an running on a single host, that doesn't involve the AIO script I've been using? perhaps running each playbook seperately? | 21:02 |
| xar- | ansible keystone, ansible nova, etc. | 21:02 |
| palendae | Yeah, I've been meaning to write a quickstart guide for just that purpose | 21:03 |
| palendae | We have our install guide for multinode deployments | 21:03 |
| palendae | 1s | 21:03 |
| palendae | http://os-ansible-deployment.readthedocs.org/en/latest/install-guide/index.html | 21:03 |
| xar- | well multinode wouldn't be suitable for a single bare metal server, or am I mistaken? | 21:03 |
| palendae | Right, but it'll give you an idea | 21:03 |
| palendae | And forthe configuration, I think you can at least use the in-tree *.aio config files | 21:04 |
| palendae | Mostly it's the configuration | 21:04 |
| palendae | the openstack_user_config.yml file | 21:04 |
| palendae | Then it's running setup-hosts.yml, setup-infrastructure.yml, setup-openstack.yml | 21:04 |
| xar- | I just rekicked via KVM(IPMI) so I've got a fresh box to keep testing on, until it works that is... ^_^ | 21:05 |
| palendae | Also, contributions are taken through http://git.openstack.org/cgit/stackforge/os-ansible-deployment - see http://git.openstack.org/cgit/stackforge/os-ansible-deployment/tree/CONTRIBUTING.rst | 21:05 |
| palendae | GitHub's basically a mirror | 21:05 |
| xar- | I think i'd be embarrassed to contribute my mediocre hacky code lol | 21:06 |
| xar- | (good to know though) | 21:06 |
| evrardjp | palendae: when you say "he is on our team now", you mean it's new or ? | 21:07 |
| palendae | He moved within the org yesterday or something | 21:07 |
| evrardjp | oh ok | 21:08 |
| evrardjp | cool :) | 21:09 |
| evrardjp | welcome mhayden then :) | 21:09 |
| mhayden | evrardjp: howdy! :) | 21:09 |
| evrardjp | you'll need to update your linked in, I just went to see it ;) | 21:10 |
| *** arbrandes has quit IRC | 21:11 | |
| mhayden | haha, that's on the very end of my very long to-do list ;) | 21:13 |
| mhayden | palendae: thanks for helping me stumble through that commit ;) | 21:14 |
| palendae | mhayden: np | 21:14 |
| odyssey4me | xar- ping | 21:16 |
| odyssey4me | mhayden ping | 21:16 |
| xar- | hey odyssey4me | 21:19 |
| xar- | im here lol | 21:19 |
| odyssey4me | palendae mhayden I've been considering what we may want to do for plugins to ansible which may or may not have an incompatible license (and we really don't want the dispute either). Essentially I'm thinking that we add a new repo in rcbops (or wherever in public), then we add scaffolding to the gate check script to pull then down and use them for the gate check. That way they don't need to be in-repo but they can still add value to the | 21:20 |
| odyssey4me | gate check. Thoughts? | 21:20 |
| odyssey4me | xar- :) | 21:21 |
| odyssey4me | alright, happy to catch you | 21:21 |
| odyssey4me | so the function for the disk space was primarily built for the gate checks which operate in rax and hpcloud | 21:21 |
| palendae | odyssey4me: I don't think it's necessary in this case; MIT's compatible and the foundation has a FAQ specifically saying what to do with MIT | 21:21 |
| odyssey4me | but we'd like to see it extended to work for vagrant and other environments as more of a general tool | 21:21 |
| xar- | yea I was terrified when I saw it do a mkpart on a block device that already held data lol | 21:22 |
| xar- | (shame on me for not reading) | 21:23 |
| odyssey4me | palendae ok, that's pretty good news... in that case I think we need to add scaffolding to make it optional rather than permanent - I don't see (personally) why this should be used every time, although I am interested in adding similar functions. I was actually inspecting this very plugin this morning to replace our timers in run-playbooks. | 21:23 |
| palendae | Cool | 21:23 |
| odyssey4me | xar- so it's not supposed to do that if you only have one disk | 21:23 |
| palendae | IMO, in general, this should definitely be on in the gate. Real deploys...questionable | 21:23 |
| xar- | in my case, I've got an SSD OS disk with free physical extents in the VG, then a large multi-terabyte (mechanical) storage array (RAID6) | 21:24 |
| odyssey4me | if you have one disk, it's supposed to use var/lib/lxc for the lxc qcow2 disks and create loopback disks for cinder-volume, swap and swift - what you may have seen is the mkpart on the swift/swap/cinder loopback disks? | 21:24 |
| xar- | it enumerated my (mdadm) array as sdb, sdc, etc. and proceeded to create new (GPT) partition tables on them | 21:25 |
| xar- | I mean, I understand its not going to be one size fits all, I'm open to writing the logic in to be more... accommodating :) | 21:25 |
| odyssey4me | xar- ah, yes - ok, let's step through this | 21:25 |
| xar- | blk_devices=$(lsblk -nrdo NAME,TYPE,RO | awk '/d[b-z]+ disk [^1]/ {print $1}') | 21:26 |
| xar- | for blk_dev in ${blk_devices}; do | 21:26 |
| odyssey4me | patches are welcome, so perhaps I can just step you through it a little and give you some guidance | 21:26 |
| xar- | that ^ :D | 21:26 |
| odyssey4me | so yes, that checks for devices other than the first | 21:26 |
| odyssey4me | and other than read only disks (cdrom, config drive, etc) | 21:27 |
| xar- | if ! vgs lxc > /dev/null 2>&1; then | 21:27 |
| xar- | parted --script /dev/${blk_dev} mklabel gpt | 21:27 |
| xar- | so yea this is the part that made me cringe, I think it would be best to add some sort of variable boolean to the above config, and have it step through an additoinal set of conditionals | 21:27 |
| odyssey4me | we do have a note in https://github.com/stackforge/os-ansible-deployment/blob/master/development-stack.rst which says that you shouldn't run this on a host you can't risk destroying :p | 21:28 |
| xar- | which is why I'm not mad haha :D | 21:28 |
| odyssey4me | hmm, so the trisk here would be perhaps to provide something that excludes drives in the regex here: https://github.com/stackforge/os-ansible-deployment/blob/master/scripts/scripts-library.sh#L72 | 21:29 |
| *** arbrandes has joined #openstack-ansible | 21:29 | |
| xar- | I think it would be preferable to have it configurable by the end user; perhaps someone other than myself would be ok with such a configuration | 21:30 |
| *** yaya has quit IRC | 21:30 | |
| evrardjp | odyssey4me: I don't agree with what you said in terms of license | 21:30 |
| *** Mudpuppy has quit IRC | 21:30 | |
| odyssey4me | or perhaps to check for a var that has the disk device that can be used - if the var isn't there, then revert to the current behaviour? | 21:30 |
| Apsu | xar-: We really just intended it to be deployed on specific flavors of specific images on specific clouds. It was a starting point :D | 21:30 |
| evrardjp | I'm not a lawyer, so it should be checked elsewhere, but I think if you include some part of the code with another incompatible license, you may have for your product to follow the license that has been included | 21:30 |
| Apsu | But yeah, blindly assuming all devices bd-bz are good to go is a bit cringe inducing | 21:30 |
| odyssey4me | xar- alternatively just have a device list which is in a var, then we set that var in the gate-check-script | 21:31 |
| evrardjp | so, incompatible licenses have REALLY to be carefully planned before integration | 21:31 |
| odyssey4me | I think I might like that idea. | 21:31 |
| Apsu | Most appropriate would probably be to specify a specific mapping, and use the default if it's not set | 21:32 |
| Apsu | Or clear guidance on what storage configuration to setup ahead of time that the AIO expects, so it'll skip making its own | 21:32 |
| odyssey4me | evrardjp yes, that's kinda why I prefer to have any non Apache2 license elsewhere so that we don't risk polluting - but the OpenStack foundation does have some notes on compatible licenses which palendae referred to | 21:32 |
| xar- | yup, for now what I've done is simply shifted it to /d[z] so that it finds nothing ^_^ | 21:32 |
| Apsu | But that feels a little like trying to make it robust, and that's dangerously close to "we support this method" | 21:33 |
| evrardjp | odyssey4me: yes, I posted the link ;) | 21:33 |
| Apsu | And that's not a thing :D | 21:33 |
| evrardjp | but it's not what I meant | 21:33 |
| evrardjp | the simple fact to include something from elsewhere could be a problem | 21:33 |
| evrardjp | (with incompatible license) | 21:33 |
| odyssey4me | xar- I like your idea - if you can pop a patch in for review then we can work on making it betterer :) | 21:33 |
| odyssey4me | evrardjp if we consume it, but don't hold it in-tree (like ansible), then there is no risk | 21:34 |
| *** KLevenstein has quit IRC | 21:34 | |
| evrardjp | http://stackoverflow.com/questions/11819831/including-gpl-external-library-in-mit-code | 21:35 |
| xar- | ok, I'll work on it :) | 21:35 |
| evrardjp | I think lawyers don't get any idea of where it's stored ;) | 21:35 |
| xar- | is there documentation on contributing to git.openstack.org? | 21:35 |
| evrardjp | odyssey4me: but I'm no expert in that matter, I'm just a cautious guy | 21:36 |
| odyssey4me | xar- we link to the good stuff from https://github.com/stackforge/os-ansible-deployment/blob/master/CONTRIBUTING.rst - check the section in 'contributing code' :) | 21:36 |
| palendae | For a higher level - https://wiki.openstack.org/wiki/How_To_Contribute | 21:36 |
| palendae | xar-: ^ | 21:36 |
| odyssey4me | if you get stuck, we're often around to help :) | 21:36 |
| evrardjp | and it's definitely easier to replace if it's an internal dependency | 21:37 |
| odyssey4me | evrardjp yeah, we've done the rounds on this with regards to GPL code and we know that it's a no-go for in-tree inclusion | 21:37 |
| palendae | GPL is specifically a no, no, yes | 21:38 |
| palendae | With Apache12 | 21:38 |
| palendae | Apache2 | 21:38 |
| bgmccollum | odyssey4me: re: MaxStartups...should a fix target the bug re-opened for trunk, or a new bug? | 21:41 |
| xar- | odyssey4me: the other thing I was fuzzy on was the cinder loopback devices, if data is written into those block devices, where exactly does it reside? | 21:41 |
| odyssey4me | bgmccollum if you have a suggested fix then use the same bug for now | 21:41 |
| odyssey4me | bgmccollum I'm not sure if we should add a code fix though, it may be better to just document a recommendation in the docs? it feels to me like the ansible forks setting should really be a deployer decision | 21:42 |
| bgmccollum | odyssey4me: yes a deployer decision, but things tend to tip over > 10 | 21:43 |
| palendae | odyssey4me: So hand people a footgun? | 21:44 |
| bgmccollum | odyssey4me: in a non-aio, you have ≈ 20 containers on an infra...so with forks left at 10, there's parallelism being left on the table | 21:44 |
| palendae | Going through severval deployments in the last coupple weeks, I can tell you people want it set | 21:44 |
| *** sdake_ has joined #openstack-ansible | 21:44 | |
| odyssey4me | bgmccollum yes, so I'm thinking that we should remove the forks setting entirely - we have no business setting it... we can have a doc note which outlines some ansible settings of interest to deployers though :) | 21:44 |
| palendae | odyssey4me: I don't think that solves anything | 21:45 |
| palendae | Feedback I've gotten is 100% people want it set | 21:45 |
| odyssey4me | bgmccollum that said, let's see a patch and discuss it in review :) | 21:45 |
| odyssey4me | palendae for the purpose of... ? | 21:45 |
| palendae | Ease | 21:45 |
| odyssey4me | are we perhaps doing too much spoon feeding I wonder? | 21:46 |
| bgmccollum | odyssey4me: its not so much fiddling the number of forks...its reconfiguring SSHD on the target hosts... | 21:46 |
| palendae | If you disagree, you are totally welcome to argue it internally | 21:46 |
| odyssey4me | palendae I'm not trying to make an argument. I'm just trying to ask an honest question. | 21:46 |
| *** sdake has quit IRC | 21:48 | |
| bgmccollum | odyssey4me: first pass can just be change to the gate scripts, and see if the runtime / retry / general failures decreases...? | 21:49 |
| *** arbrandes has quit IRC | 21:50 | |
| *** pradk has quit IRC | 21:53 | |
| *** KLevenstein has joined #openstack-ansible | 21:54 | |
| openstackgerrit | Steve Lewis proposed stackforge/os-ansible-deployment: Add support for swift proxy config: sorting_method https://review.openstack.org/208817 | 21:55 |
| odyssey4me | bgmccollum so right now in the gate we don't have retry issues any more - the issues are only showing up in larger deployments and when using proper hardware for testing | 21:57 |
| bgmccollum | odyssey4me: ill do some testing with some onmetal servers and report back | 21:58 |
| odyssey4me | bgmccollum great, thank you for the research and help! | 22:01 |
| *** sdake_ is now known as sdake | 22:01 | |
| stevelle | if I want two roles to have the same handler would we rather have the handler duplicated or something else? | 22:03 |
| *** arbrandes has joined #openstack-ansible | 22:06 | |
| *** KLevenstein has quit IRC | 22:13 | |
| openstackgerrit | Neill Cox proposed stackforge/os-ansible-deployment: Stop configure_diskspace from using small partitions https://review.openstack.org/216880 | 22:16 |
| *** sdake_ has joined #openstack-ansible | 22:21 | |
| *** sdake has quit IRC | 22:24 | |
| *** sdake_ is now known as sdake | 22:24 | |
| odyssey4me | stevelle ansible doesn't to DRY well - duplicate it :) | 22:25 |
| stevelle | yeah, always afraid of that | 22:25 |
| palendae | You can specifically include a handlers/main.yml file | 22:25 |
| palendae | But if it crosses roles, that's bad | 22:25 |
| *** abitha has quit IRC | 22:28 | |
| *** spotz is now known as spotz_zzz | 22:52 | |
| *** KLevenstein has joined #openstack-ansible | 22:53 | |
| *** cloudtrainme has quit IRC | 22:53 | |
| odyssey4me | xar- you still around? | 23:00 |
| *** shoutm has joined #openstack-ansible | 23:01 | |
| openstackgerrit | Jesse Pretorius proposed stackforge/os-ansible-deployment: Add profiling for Ansible tasks https://review.openstack.org/216849 | 23:10 |
| *** daneyon has quit IRC | 23:18 | |
| *** scarlisle has quit IRC | 23:25 | |
| *** bitblt has joined #openstack-ansible | 23:28 | |
| *** bitblt has quit IRC | 23:28 | |
| *** misc has quit IRC | 23:33 | |
| *** KLevenstein has quit IRC | 23:33 | |
| *** misc has joined #openstack-ansible | 23:34 | |
| xar- | odyssey4me: yup :D | 23:38 |
| xar- | im always around, znc! :) | 23:39 |
| *** arbrandes has quit IRC | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
