15:00:25 <jpena> #startmeeting RDO meeting (2016-06-15)
15:00:26 <zodbot> Meeting started Wed Jun 15 15:00:25 2016 UTC.  The chair is jpena. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:26 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:26 <zodbot> The meeting name has been set to 'rdo_meeting_(2016-06-15)'
15:00:26 <openstack> Meeting started Wed Jun 15 15:00:25 2016 UTC and is due to finish in 60 minutes.  The chair is jpena. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:27 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:30 <openstack> The meeting name has been set to 'rdo_meeting__2016_06_15_'
15:00:46 <trown|mtg> o/
15:00:49 <chandankumar> \o/
15:00:55 <weshay> ellooo
15:00:57 <jpena> #topic roll call
15:01:01 <dmsimard> o/
15:01:02 <rbowen> /o/
15:01:09 <amoralej> o/
15:01:14 <hrybacki> o/
15:01:25 <jpena> #chair imcsk8 trown|mtg chandankumar weshay dmsimard rbowen amoralej hrybacki
15:01:25 <zodbot> Current chairs: amoralej chandankumar dmsimard hrybacki imcsk8 jpena rbowen trown|mtg weshay
15:01:26 <openstack> Current chairs: amoralej chandankumar dmsimard hrybacki imcsk8 jpena rbowen trown|mtg weshay
15:01:33 <rbowen> ETOOMANYBOTS
15:01:34 <panda> ooo/
15:01:47 <jpena> dmsimard: can you do your magic?
15:02:09 <jpena> #chair panda number80
15:02:09 <zodbot> Current chairs: amoralej chandankumar dmsimard hrybacki imcsk8 jpena number80 panda rbowen trown|mtg weshay
15:02:10 <openstack> Current chairs: amoralej chandankumar dmsimard hrybacki imcsk8 jpena number80 panda rbowen trown|mtg weshay
15:02:12 <number80> I tried
15:02:17 <number80> seems working
15:02:20 <dmsimard> yeah
15:02:25 <dmsimard> openstack is permanently +q'd
15:02:28 <jpena> ok, let's start
15:02:28 <dmsimard> so without +v he can't talk :)
15:02:38 <jpena> #topic DLRN instance migration to ci.centos infra
15:02:49 <jpena> we're pretty close, really :)
15:03:25 <dmsimard> jpena: did the internal dlrn catch up on the build failures ?
15:03:36 <jpena> dmsimard: it did
15:04:05 <dmsimard> so I guess the only thing missing now is the consistent symlink
15:04:11 <jpena> so the only pending topic before performing the switch is getting the consistent link synced, so the puppet promotion scripts are happy
15:04:24 <dmsimard> always EmilienM's fault
15:04:26 <jpena> that's https://review.rdoproject.org/r/1407
15:04:47 <EmilienM> lol
15:04:53 <jpena> once it's merged, we'll get it working on the new instance, test it, then do the cut
15:04:57 <dmsimard> jpena: wait
15:04:58 <dmsimard> that review
15:05:24 <dmsimard> that means the current symlink is not symlink'd ?
15:05:36 <dmsimard> sensu monitors ftbfs through current/versions.csv
15:05:53 <dmsimard> er, current symlink is not rsync'd*
15:06:06 <jpena> dmsimard: ooookey, so we might need to symlink current as well :/
15:06:11 <jpena> apevec: ^^
15:06:15 <dmsimard> well, unless you have another idea
15:06:28 <jpena> where is the sensu server?
15:06:33 <dmsimard> https://github.com/rdo-infra/ansible-role-rdomonitoring/blob/master/files/check-delorean-builds.py
15:06:38 * apevec is late
15:06:41 <jpena> if it's inside the ci.centos infra, it could use trunk-primary
15:06:44 <dmsimard> jpena: oh, you have a good point
15:06:45 <Duck> quack
15:06:53 <jpena> #chair apevec Duck
15:06:53 <zodbot> Current chairs: Duck amoralej apevec chandankumar dmsimard hrybacki imcsk8 jpena number80 panda rbowen trown|mtg weshay
15:06:54 <openstack> Current chairs: Duck amoralej apevec chandankumar dmsimard hrybacki imcsk8 jpena number80 panda rbowen trown|mtg weshay
15:06:59 <dmsimard> jpena: right now the ftbfs is monitored from the master, but it could be monitored from the dlrn server itself
15:07:04 <Duck> thanks, I was a bit tired :-)
15:07:09 <dmsimard> nevermind, I'll adapt the monitoring
15:07:30 <jpena> dmsimard: ok, so we can keep on without current?
15:07:37 <dmsimard> yeah I don't need it
15:07:43 <jpena> cool
15:08:02 <dmsimard> sorry for the panic :)
15:08:23 <jpena> EmilienM: btw, I see puppet-openstack-integration uses a specific hashed repo. It's present in the new server, but maybe some package is not the same due to different build order
15:08:47 <dmsimard> yeah, the hashes on the new instance *can* be different
15:08:47 <apevec> jpena, what do you mean?
15:08:51 <dmsimard> there might be a period of limbo
15:08:52 <EmilienM> jpena: we have a periodic (daily) that update our manifest with latest URL
15:09:10 <jpena> apevec: the current hash is in  https://github.com/openstack/puppet-openstack-integration/blob/master/manifests/repos.pp#L27
15:09:11 <dmsimard> apevec: tl;dr the rcip instance lags behind and can have a different dist git hash for the same upstream commit hash
15:09:13 <apevec> jpena, ah diff now during transition?
15:09:18 <jpena> yep
15:09:19 <EmilienM> see https://review.openstack.org/#/c/328591/
15:09:34 <jpena> it's just a warning, we should be ready for hiccups after migration
15:09:59 <apevec> yeah
15:10:13 <jpena> so action
15:10:24 <apevec> but as soon as puppet promoption succeeds against new trunk.rdo, we should be good
15:10:47 <jpena> #action jpena to coordinate dlrn server switch once https://review.rdoproject.org/r/1407 is merged and tested in the new server
15:10:55 <dmsimard> jpena: emphasis on tested
15:10:58 <dmsimard> rsyncing symlinks is weird sometimes
15:11:23 <jpena> I think I tested it when I first prepared the patch, then removed it
15:11:28 <jpena> anyway, we'll test it for sure
15:11:42 <jpena> anything else?
15:12:17 <apevec> "to coordinate" includes it all :)
15:12:35 <jpena> #topic openstack/packstack stable/kilo EOL
15:12:58 <apevec> I've put that, can we just eol it
15:13:06 <dmsimard> why just packstack ?
15:13:09 <apevec> or anyone seems some work we want to push it?
15:13:13 <apevec> dmsimard, it is not just
15:13:23 <apevec> most of projects have been eoled
15:13:27 <apevec> see that thread
15:13:36 <dmsimard> yeah, nevermind, I read that as "github.com/openstack/packstack" :)
15:13:39 <apevec> this is about cleaning up remaining stable/kilo
15:14:09 <Duck> number80: quack
15:14:25 <imcsk8> apevec: what's the EOL process?  just stop creating packages?
15:14:31 <apevec> jayg, there's also OPM stable/kilo
15:14:52 <number80> process is to stop syncing repo and blocking builds from koji
15:14:53 <apevec> imcsk8, upstream EOL means pushing eol-kilo tag and deleting the branch
15:15:02 <jpena> there are no pending reviews to packstack in kilo
15:15:16 <jayg> apevec: that may be more risky to drop, though borderline
15:15:33 <apevec> number80, ah right, we also have work on rpmfactory side
15:15:37 <imcsk8> apevec: and setting bugs to EOL right?
15:15:50 <jayg> I don't think packstack ever used it, and the usage in ofi/staypuft was fairly minimal, but it was around juno or kilo I think
15:16:04 <apevec> imcsk8, yes, and open gerrit reviews need to be closed for that branch
15:16:34 <number80> apevec: I still have stuff in kilo queue so not yet :)
15:16:37 <imcsk8> i'll check for lauchpad bugs, the rhbz ones are covered with the script i think...
15:16:57 <apevec> number80, right, topic is only about upstream packstack repo
15:17:21 <apevec> I'll bring up rdo kilo EOL process in few weeks
15:17:44 <chandankumar> imcsk8, apevec https://github.com/openstack-infra/release-tools/blob/master/expire_old_bug_reports.py
15:17:58 <apevec> chandankumar, this is about LP
15:18:03 <jayg> ohh, you are not talking about the certmonger change I guess, sorry, I hadn't read chat before ping
15:18:28 <chandankumar> apevec, yes
15:18:29 <apevec> imcsk8, ack - let's coordinate then
15:18:38 <imcsk8> apevec: the script addresses launchpad
15:18:53 <imcsk8> chandankumar: thanks!
15:19:14 <apevec> #action apevec and imcsk8 to coordinate openstack/packstack eol-kilo after checking upstream Launchpad bugs
15:19:57 <apevec> chandankumar, imcsk8 ah ok, but I  don't think we want to just expire based on time
15:20:17 <imcsk8> should be based on version
15:20:30 <chandankumar> imcsk8, yes,
15:20:32 <apevec> imcsk8, can you try that script in dry mode first and share the list of bugs?
15:20:42 <imcsk8> apevec: yes
15:20:54 <apevec> anyway, there are details we can take off-meeting
15:20:58 <apevec> next topic
15:21:21 <jpena> #topic graylist review.rdoproject.org
15:22:07 <apevec> that's still not solved,
15:22:16 <apevec> but there's progress update in https://trello.com/c/vgSUYFjU/162-email-graylist-review-rdoproject-org
15:22:25 <apevec> jschlueter, fbo ^
15:23:29 <fbo> Yes I think the main issue is the reverse DNS configuration like I said in the trello ticket
15:24:15 <apevec> ok, thanks
15:24:34 <apevec> we'll keep it on the agenda and check next week
15:24:39 <fbo> We have the same issue on sf-project.io, I asked the helpdesk to fix the reverse for sf-project.io and then I'll validate if it work as expected then we can do the same for review.rdoproject.org
15:25:05 <fbo> But I need the reverse to be fixed before investigating more
15:25:08 <Dog> Duck: woof
15:25:34 <apevec> Duck, ^ can you help speed up reverse dns fix?
15:25:54 <apevec> or misc, with your IT connections? :)
15:26:20 <misc> apevec: mhh a ticket was opened ?
15:26:28 <apevec> fbo opened it
15:26:32 <fbo> Yes I have no response so far from the helpdesk. Let me check misc
15:27:48 <fbo> misc, REQ0706907
15:28:09 <fbo> or RITM0162532 don't know which one you need
15:28:33 <misc> I can find the 2nd from the first
15:28:56 * Duck pat the Dog
15:29:03 <misc> fbo: now, I do not see why the reverse DNS is related to greylisting
15:29:37 <rbowen> Some MTAs reject email if they can't reverse lookup the source, no?
15:29:46 <Duck> apevec: I don't have connections :-)
15:29:54 <fbo> misc, the smtp server present itself with the domain but the reverse is different
15:29:54 <misc> sure, some do, but then, that's not greylisting, that's being rejected :)
15:30:17 <Duck> tep
15:30:20 <Duck> yep
15:30:28 <misc> fbo: and 133.231.46.in-addr.arpa is managed by nse2.enovance.com
15:30:51 <misc> not sure if RH IT manage this one
15:31:00 <Duck> if the direct of the reverse is not identical we can expect a lot of reject
15:31:04 <fbo> misc, I don't know
15:31:13 <jschlueter> misc: milter-graylist was complaining that helo=XXX didn't resolve to anything thing
15:32:18 <misc> fbo: that's a server in equinix2, so likely not RH IT
15:32:30 <dmsimard> are we going to troubleshoot this in the middle of the meeting ?
15:32:45 <jpena> yes, that's what I was thinking. next topic?
15:32:48 <misc> dmsimard: why not :p but yeah, we cna discuss later
15:33:00 <fbo> dmsimard, no, I'll check with IT guys from enovance and keep you updated
15:33:33 <number80> please add action ;)
15:33:48 <Duck> fbo: could you do so on the ML, so we can follow?
15:33:53 <number80> +> #action <who> <what>
15:34:10 <fbo> #action fbo followup with reverse dns issue
15:34:24 <number80> thanks :)
15:34:33 <jpena> #topic notify trello board changes in #rdo?
15:34:53 <jpena> who proposed this?
15:34:55 <chandankumar> +1 to that
15:34:57 <number80> this is just my opinion: yes but moderately
15:35:04 <dmsimard> I'm not particularly a fan of this, maybe specific actions only
15:35:10 <dmsimard> it can get spammy when someone gets trigger happy
15:35:14 <dmsimard> I've seen it in other channels
15:35:34 <chandankumar> when a card is done or new card is created notification would be good
15:35:41 <jpena> agreed. Maybe card creation and archival, or something similar
15:35:42 <jschlueter> there is already a lot of bot chatter on here ... all for limited notification only
15:36:21 <number80> if possible, I'd like to consolidate the bots as much as possible
15:36:28 <imcsk8> +1
15:36:50 <number80> (currently, I count 5, if we could go down to 3, it'd be great)
15:36:57 <dmsimard> not sure that's possible
15:37:02 <dmsimard> openstack bots are what they are
15:37:14 <number80> dmsimard: this is best effort, not requirement
15:37:22 <dmsimard> rdogerrit is openstackgerrit but for review.rdoproject (gerritbot)
15:37:43 <dmsimard> rdobot is all purpose (though only sensu mostly)
15:38:05 <dmsimard> I hate bot spam as much as the next guy
15:38:13 <number80> dmsimard: supybot as a plugin that basically listen to a socket to forward messages from other bots or scripts, we can leverage that maybe
15:38:26 <dmsimard> if a ftbfs is being worked on, someone should acknowledge the sensu alarm
15:38:29 <dmsimard> I do it sometimes
15:38:46 <dmsimard> the rest of the sensu alarms are sent to a secret channel :P
15:38:50 <dmsimard> like cpu and stuff.
15:39:00 <apevec> #rdo-dev ?
15:39:02 <dmsimard> so #rdo is mostly spared
15:39:04 <dmsimard> yeah
15:39:25 <number80> wfm
15:39:35 <dmsimard> maybe a proper infra-ish channel is warranted
15:39:42 * number80 must leave
15:39:51 <dmsimard> but fragmenting channels is not awesome either
15:40:02 <apevec> dmsimard, or separate emailing list for notifications?
15:40:25 <dmsimard> I already receive way too many emails and am subscribed to already way too many mailing lists
15:40:26 <dmsimard> personally
15:40:28 <apevec> yeah, please no separate channels for discussions
15:40:40 <Duck> btw, if there an infra ML ?
15:40:43 <apevec> dmsimard, filter :)
15:40:50 <dmsimard> I have filters
15:41:07 <apevec> Duck, no, currently we only have  rdo-list
15:41:16 <dmsimard> if it gets filtered and never read, it's not worth receiving in the first place so I end up strolling through way too much stuff
15:41:36 <dmsimard> we *could* send alarms to a mailing list, though
15:41:48 <Duck> ho, because there's a lot of things I'm useless for and I don't even understand well, so I've got a hard time following the main list
15:41:53 <dmsimard> not that I'd subscribe to it :p
15:42:17 <trown|mtg> just an FYI, there is a separate channel for tripleo-quickstart now #oooq, but that is specifically for oooq dev discussion that can really dominate this channel if kept here
15:42:21 <Duck> as misc and I are involved in many projects it is a bit difficult to follow all of this
15:43:20 <jpena> so...
15:43:33 <apevec> yeah, about the topic: yes/no ?
15:43:51 <jpena> +1 for me, provided it's limited to card creation/archival
15:43:58 <apevec> I'd say no for now on #rdo
15:44:06 <apevec> but let's experiment on #rdo-dev
15:44:10 <amoralej> even only for creation
15:44:16 <trown|mtg> +1 to experimenting on #rdo-dev
15:44:23 <trown|mtg> then it is opt-in
15:44:41 <dmsimard> rdo-dev for trello ?
15:44:43 <dmsimard> sure
15:44:48 <jpena> ok, experimenting looks like a good idea
15:45:03 <apevec> now, who knows who to set that up?
15:45:17 <apevec> is there yet another bot?
15:45:37 <trown|mtg> I think there is a trobot
15:45:47 <apevec> heh, I suspected that :)
15:45:48 * trown|mtg didnt name it
15:45:56 <rbowen> Not to be confused with trownbot
15:46:04 <imcsk8> hahaha
15:47:55 <apevec> number80 proposed this topic?
15:48:10 <dmsimard> yeah but he's gone afaik
15:48:11 <apevec> number80 ^ can you investigate trobot?
15:48:18 <apevec> ah perfect, give him action :)
15:48:41 <apevec> #action number80 to investigate trobot on #rdo-dev
15:48:44 <apevec> next topic
15:48:54 <jpena> #topic     Doc Day - https://etherpad.openstack.org/p/rdo-doc-day - June 16, 17
15:49:00 <rbowen> So, most of the information is therein the agenda.
15:49:18 <rbowen> pkovar and I will be working on the RDO website, and we would appreciate your help
15:49:31 <rbowen> in identifying outdated pages, and brainstorming what docs need written.
15:49:54 <rbowen> We also have Big Plans to update the front page to more clearly identify the various audiences, and guide them to the info that they care about.
15:50:05 <rbowen> This involved designer-fu, so is delayed until that's available.
15:50:26 <rbowen> But anyways, that's pretty much all. Join us tomorrow and Friday and help us make the website not suck.
15:50:29 <rbowen> Thanks.
15:50:35 <pkovar> +1
15:50:38 <imcsk8> count with me :)
15:50:58 <jpena> #action everyone to join the RDO Doc Day
15:51:21 <jpena> #topic Chair for next meeting
15:51:57 <imcsk8> i can do it
15:52:09 <jpena> cool, thanks imcsk8
15:52:17 <jpena> #action imcsk8 to chair next meeting
15:52:22 <imcsk8> np :)
15:52:24 <jpena> and finally
15:52:27 <jpena> #topic open floor
15:52:28 <trown> thanks imcsk8
15:53:02 <Duck> have an infra ML!!!
15:53:49 <jpena> does anyone want to add anything? If not, we can finish for today
15:53:55 <dmsimard> Duck: I think we can use rdo-list
15:54:02 <dmsimard> with an [infra] prefix if anything
15:54:17 <dmsimard> I don't think we have enough traffic to warrant an infra exclusive list
15:54:24 <Duck> I wanted to work on the Mailman 3 with number80, but I did not heard any news about it
15:54:41 <jpena> yep, let's use rdo-list for now, and migrate if needed
15:54:58 <rbowen> We should use PonyMail once we're on MM3. http://ponymail.incubator.apache.org/
15:55:01 <Duck> as I said YOU read all, and I have no much to read
15:55:20 <Duck> so I'll probably keep on not having much attention until specifically pingued
15:55:53 <jpena> Duck: would it work if we used a specific tag, as dmsimard suggested? That's filter-friendly
15:56:02 <Duck> yes it would
15:56:47 <Duck> let's find one nice and advertise on the ML then
15:57:10 <jpena> [rdo-infra] maybe?
15:58:41 <Duck> jpena: we can check the Hyperkitty interface work and test PM once the setup is ok I guess
15:58:50 <rbowen> Duck: PonyMail!
15:58:55 <rbowen> :)
15:59:04 <Duck> I read there's a MM3 plugin
15:59:31 <rbowen> For PonyMail? Yes, there is. Or you just subscribe it to the list, and it supports anything.
16:00:09 <Duck> jpena: ok for me with the tag
16:00:14 <dmsimard> ok, I have to head out
16:00:17 <dmsimard> thanks for the meeting
16:00:18 <jpena> ok, let's try it then
16:00:20 <dmsimard> food!
16:00:24 <rbowen> Also, #ponymail is a good place for discussion.
16:00:27 <jpena> running out of time, anything else?
16:00:33 <rbowen> Thanks for the meeting, folks.
16:00:47 <jpena> #endmeeting