15:00:09 <mmedvede> #startmeeting third-party
15:00:09 <openstack> Meeting started Mon Jan 25 15:00:09 2016 UTC and is due to finish in 60 minutes.  The chair is mmedvede. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:10 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:13 <openstack> The meeting name has been set to 'third_party'
15:00:41 <mmedvede> hi third-party CI folks. Anyone here for the meeting?
15:00:46 <lennyb> hi
15:00:52 <cbader> hi
15:00:54 <Swanson> Hello
15:00:55 <asselin__> good morning
15:00:58 <mmedvede> I am sitting in today for anteaya
15:01:33 <mmedvede> hi all, good morning
15:02:14 <mmedvede> I have no announcements to make. Anyone has a topic they would like to discuss?
15:02:57 <Swanson> Not sure what is normally done at these meetings.
15:03:25 <cbader> this is my first meeting so don't know either
15:03:42 <mmedvede> Swanson: normally we answer questions that anyone who is running third-party CI has, and helping with problems they have
15:03:48 <mmedvede> cbader: ^
15:03:52 <asselin__> usually people bring up issues they have or recently solved
15:04:05 <Swanson> I've been having trouble getting jenkins to log into newly created nodes.  I can use the keys and whatnot by hand to log in but jenkins can't seem to do this.
15:04:09 <Swanson> Anyone seen that?
15:04:39 <asselin__> cbader, looks like your issue right? ^^
15:04:40 <cbader> You might want to try to add the private key instead of pointing it to the file
15:04:52 <lennyb> I saw something similar, re adding nodes to Jenkins solved it, I think
15:05:04 <cbader> yes I have not figured out why it doesn't use the file yet.
15:05:25 <mmedvede> I also use key directly, not pointing to file
15:05:50 <lennyb> we are using user/psw in our env
15:06:24 <cbader> I am still having a issue with zuul not connecting to gerrit. it is complaining about credentials but I have verified the private/public key and user.
15:06:39 <Swanson> Do you have that happening automagically on the node creation?  Cause once the node is recreated I have to go back in and do it again.
15:07:55 <cbader> Are you going to Jenkins>Manage Credentials> Private Key (Enter directly) that was the only thing the worked for me.
15:08:19 <lennyb> cbader: trivial question: is this gerrit account enabled?
15:08:31 <mmedvede> cbader: did you try start zuul in debug mode? 'su zuul; zuul-server -d'?
15:08:32 <asselin__> all: there's an etherpad now to try to track issues and experiences: https://etherpad.openstack.org/p/thirdpartyci
15:08:54 <asselin__> #link 3rd party ci - setup etherpad https://etherpad.openstack.org/p/thirdpartyci
15:09:03 <Swanson> cbader, If I did that it worked.  When that node is blown up overnight I then have to do it again with the new node.
15:09:18 <asselin__> link is also saved here under "reference links" https://wiki.openstack.org/wiki/Meetings/ThirdParty#Reference_links
15:09:29 <asselin__> thats the 3rd party ci wiki homepage ^^
15:10:44 <asselin__> Swanson, so you did that and it only works when you manually register a jenkins slave with a jenkins amster?
15:10:50 <cbader> I will look for my issue there thanks
15:11:48 <Swanson> asselin__, Mostly.  Sometimes adding a key into jenkins doesn't work either.
15:12:03 <Swanson> asselin__, but I haven't looked into what was up with that too closely.
15:13:03 <asselin__> Swanson, just note that puppet probably removes the key and restores the link to the private key filename...
15:13:35 <asselin__> Swanson, http://git.openstack.org/cgit/openstack-infra/puppet-openstackci/tree/templates/be.certipost.hudson.plugin.SCPRepositoryPublisher.xml.erb#n10
15:13:54 <asselin__> actually....that's for scp....
15:14:34 <Swanson> asselin__, And certainly when the node is replaced what I did for the previous node is irrelevant
15:14:46 <Swanson> VPN she be a harsh mistress today.
15:15:40 <asselin__> Swanson, so you have secured jenkins?
15:15:50 <mmedvede> Swanson: and your nodes, nodepool does update them daily?
15:17:14 <Swanson> asselin__, I have not secured jenkins.  Wide open.  Should I?
15:17:33 <Swanson> mmedvede, nodepool updates them daily.
15:17:35 <asselin__> Swanson, no...it should be optional....
15:20:32 <mmedvede> anything else on this topic?
15:21:11 <Swanson> I'll add this to the thirdpartyci document once it chooses to load.
15:21:25 <Swanson> I think it had already been listed.
15:21:32 <cbader> So looks like pip 8.0.2 is out are people using it did it fix the issues from last week?
15:23:35 <mmedvede> cbader: can you refresh us on the issues from last week?
15:24:12 <mmedvede> We have not seen any troubles in our CI
15:24:50 <cbader> there were issues with argparse was causing issues with modules not being able to load because it pip couldn't read what the requirements were needed.
15:26:02 <cbader> I was creating a new Common-CI and ran into this issues so pinned mine to 7.1.2 to continue.
15:26:36 <lennyb> cbader: what version were you using before?
15:28:01 <cbader> lennyb, It was a new install so I got what it pulled. Which was 8.0.0 so with what was being talked about I backed up to the last working version 7.1.2 and my issue went away.
15:31:05 <Swanson> In this meeting lull I would just like to throw out there my apologies to the world for my rouge ci spamming last week.  At least I caught it quickly.
15:33:37 <mmedvede> Swanson: I am interested to hear what caused it. That is my biggest fear :)
15:34:25 <Swanson> I screwed up and swapped out my small list of jenkins jobs for what I assume is the gigantic list of Infras jenkins jobs.
15:35:42 <Swanson> "I don't remember this file being this big...  Sure am getting a lot of email..."
15:36:20 <mmedvede> haha, that would do it
15:37:21 <mmedvede> I guess you caught it fast, because I did not see any email on -dev or -infra lists about it
15:38:20 <mmedvede> Swanson: thank you for sharing
15:38:40 <Swanson> mmedvede, I happy to serve as a warning to others.
15:39:15 <asselin__> Swanson, sorry...I didn't catch what was the root cause
15:40:20 <Swanson> asselin__, Updating some of my local project config files I managed to overwrite some files I didn't want to overwrite.  (jenkins jobs for one)  So be careful and double check your commits before starting up your cI.
15:41:12 <asselin__> Swanson, which files exactly?
15:42:28 <Swanson> asselin__, projects.yaml, I believe.
15:43:00 <asselin__> hmmm....I don't know why any change to that would cause out of control comments....
15:43:12 <mmedvede> Swanson: oh, I assumed it was zuul config
15:43:23 <asselin__> Swanson, do you have tox in your project-config repo?
15:43:33 <mmedvede> not sure how changes just to jenkins config would make your CI spam
15:43:35 <Swanson> asselin__, I overwrote it with infra's.
15:44:21 <mmedvede> Swanson: do you use zuul, or jenkins gerrit plugin?
15:44:39 <Swanson> I'm sorry.  I also overwrote zuul's layout.yaml
15:45:02 <Swanson> I did both.  I caught jobs and then got nailed by zuul.  Same reason for both.
15:46:14 <Swanson> Updating elements and scripts and whatnot I ended up updating everything.
15:47:09 <asselin__> ok I see...you were re-syncing your project config nodepool files with upstreams project-config and copied more files than you intended
15:47:51 <Swanson> asselin__, That would be it exactly and with the clarity I should have provided in the first place.
15:48:05 <asselin__> ok :)
15:48:40 <asselin__> so perhaps a warning would be helpful?
15:50:09 <Swanson> It wouldn't hurt.  The thing is when you're doing things fast and irritated things like this happen.  A reminder that you could be yelled at by anteaya (sp?) wouldn't hurt.
15:53:39 <mmedvede> fair warning on zuul config might be good
15:53:47 <mmedvede> anything else?
15:55:15 <mmedvede> let's get a few minutes back then
15:55:19 <mmedvede> good discussion today, thank you everyone
15:55:24 <Swanson> Thanks
15:55:29 <cbader> thanks
15:55:44 <mmedvede> #endmeeting