Friday, 2017-07-28

*** thorst has quit IRC00:00
*** edmondsw has joined #openstack-powervm01:25
*** edmondsw has quit IRC01:30
*** miltonm has quit IRC01:31
*** esberglu has quit IRC01:42
*** thorst has joined #openstack-powervm01:57
*** thorst has quit IRC02:02
*** esberglu has joined #openstack-powervm02:11
*** esberglu has quit IRC02:16
*** edmondsw has joined #openstack-powervm03:13
*** edmondsw has quit IRC03:18
*** thorst has joined #openstack-powervm03:23
*** thorst has quit IRC03:37
*** edmondsw has joined #openstack-powervm05:02
*** edmondsw has quit IRC05:06
*** thorst has joined #openstack-powervm06:00
*** thorst has quit IRC06:06
*** esberglu has joined #openstack-powervm06:31
*** esberglu has quit IRC06:36
*** edmondsw has joined #openstack-powervm06:50
*** edmondsw has quit IRC06:54
*** thorst has joined #openstack-powervm08:03
*** thorst has quit IRC08:07
*** openstackgerrit has quit IRC08:18
*** esberglu has joined #openstack-powervm08:19
*** esberglu has quit IRC08:23
*** thorst has joined #openstack-powervm10:04
*** esberglu has joined #openstack-powervm10:07
*** thorst has quit IRC10:09
*** esberglu has quit IRC10:12
*** edmondsw has joined #openstack-powervm10:26
*** edmondsw has quit IRC10:30
*** thorst has joined #openstack-powervm11:40
*** AlexeyAbashkin has joined #openstack-powervm11:40
*** thorst has quit IRC11:44
*** AlexeyAbashkin has quit IRC11:47
*** esberglu has joined #openstack-powervm11:55
*** esberglu has quit IRC12:00
*** edmondsw has joined #openstack-powervm12:13
*** thorst has joined #openstack-powervm12:16
*** smatzek has joined #openstack-powervm12:28
*** apearson has joined #openstack-powervm12:52
*** esberglu has joined #openstack-powervm12:55
*** esberglu has quit IRC12:55
*** esberglu has joined #openstack-powervm12:56
*** esberglu has quit IRC13:01
*** esberglu has joined #openstack-powervm13:16
*** AlexeyAbashkin has joined #openstack-powervm14:19
*** kjw3 has quit IRC14:32
*** efried is now known as fried_rice14:36
*** edmondsw has quit IRC15:06
*** edmondsw has joined #openstack-powervm15:19
esberglufried_rice: edmondsw: By having remote pypowervm set up before stacking I think we can also get rid of the discover_hosts loop post stack.15:29
esbergluSetting up the environment files beforehand allowed stacking with that patch that broke us15:29
fried_riceesberglu Sweet.15:30
esberglufried_rice: edmondsw: 5617 for the above. Still need to put it through a full CI run though15:48
edmondswesberglu commented15:56
*** apearson has quit IRC16:00
*** thorst_ has joined #openstack-powervm16:02
*** apearson has joined #openstack-powervm16:02
*** apearson has quit IRC16:03
*** apearson has joined #openstack-powervm16:03
*** thorst has quit IRC16:05
esbergluedmondsw: Replied16:07
edmondswesberglu couldn't you sudo the touch and iniset, and then remove both chowns?16:09
edmondswif not, don't you need to chown the files as well as the dir back to root at the end?16:10
esbergluedmondsw: Yeah I added a second comment just now with chown fix. Nope sudo touch/iniset doesn't work.16:11
esbergluThe sudo iniset fails16:11
esbergluI think because the source of iniset doesn't apply to sudo16:11
esbergluI was getting an iniset command not found when I tried that16:11
esbergluedmondsw: Perhaps you know a way to get that source to work with sudo? I didn't find anything with a quick google search16:13
edmondswsudo source /opt/stack/devstack/inc/ini-config && iniset ... ?16:14
esberglusudo source doesn't work16:14
esberglusudo: source: command not found16:14
edmondswsudo . /opt/stack/devstack/inc/ini-config && iniset ... ?16:14
edmondswmay have the same problem16:14
esberglusudo . /opt/stack/devstack/inc/ini-config16:15
edmondswI never use source, just .16:15
esbergluYeah same issue. Are you cool with just chown to jenkins as long as I fix the chown back to root?16:15
edmondswnot sure what the underlying mechanism is, but . saves me 5 characters :)16:15
*** esberglu is now known as esberglu_lunch16:21
fried_riceI thought iniset takes a -sudo arg16:31
fried_riceesberglu_lunch edmondsw ^16:32
fried_riceedmondsw source and . are the same.16:32
edmondswthat's what I figured16:33
edmondswone is just faster to type :)16:33
*** openstackstatus has quit IRC16:41
*** openstack has joined #openstack-powervm16:44
thorst_fried_rice edmondsw: You should read back on #openstack-meeting.  Looks like ceilometer-powervm has some issues16:51
fried_ricethorst_ The releaseteam meeting?16:51
thorst_I just saw we were tagged...reading through it a bit more16:52
fried_riceYup, I'll fup in -releases with those guys.16:53
fried_rice-release rather16:53
thorst_muchos gracias16:54
thorst_fried_rice: maybe the question for ceilometer is should we now go into the main ceilometer branch?  They didn't want us doing that initially because they wanted decomposition.  Maybe that changed.  But I think that this is just a 'its not part of the tent' thing instead.16:56
fried_riceJoin us in -release16:57
fried_riceI don't know anything about the history.16:57
fried_riceAnd precious little about the project, for that matter.16:57
fried_riceI just know dhellman, ttx, and smcginnis ;-)16:57
thorst_let me add the channel...sec16:57
thorst_eavesdrop hasn't caught up yet...16:58
fried_riceesberglu_lunch It sounds like we were supposed to be tagging our releases of ceilometer-powervm via the official openstack release process (
fried_riceApparently that one is special because it's an official project (whereas the other -powervms aren't).17:05
thorst_so a bit of background here.17:05
fried_riceyes please17:05
thorst_when we were first doing the ceilometer bits for powervm we asked to be part of ceilometer proper17:06
thorst_after some back and forths, the ceilometer team (gordc really at the time) encouraged us to be a decomposed project, as that was what they wanted at the time17:06
thorst_it would let us be our own +2's, we'd stick to the API, etc...17:06
thorst_similar to the neutron model.17:06
thorst_but I think ceilometer-powervm is under the tent, where as neutron doesn't do that for decomposed projects17:06
thorst_and that's why I think it needs to be adhered to with the release model more closely17:07
thorst_but to be fair, I think all the -powervm projects should do that.17:07
thorst_edmondsw: ^^ when you get time, your input would be good too.17:07
fried_riceI'm looking at the history of ceilometer-powervm, and since newton, as far as I can see, there's only been one change set that did anything with real function.17:08
fried_riceAll the rest (and even arguably that one) we would have gotten for free if we were part of ceilometer proper.17:08
fried_riceshoot, even as far back as mitaka.17:09
thorst_right, it's been a super stable project17:10
thorst_don't take lack of change as a bad thing17:10
fried_riceIOW, whatever benefit we've been getting from being separate is probably diminished to zero at this point.17:10
fried_riceOh, definitely.17:10
thorst_fried_rice: well, the only challenge I suppose is that we do at some point want to get SR-IOV ism's in there17:11
thorst_but that's always been lower priority.17:11
fried_riceAnd what's the cost?  Having to have the ceilometer cores do our +2s/+As?  I think I'd be okay with that (to be fair, not having seen how much volume they have or how quickly they approve stuff).17:12
thorst_fried_rice: yeah.17:12
fried_riceWe would have to twiddle the CI, but having done nova in-tree, I'm confident that's doable.17:12
*** esberglu_lunch is now known as esberglu17:12
-openstackstatus- NOTICE: The Gerrit service on is being taken offline for roughly 5 minutes to perform a database backup and reconfiguration17:13
esbergluthorst_: fried_rice: Yeah that would be super easy from a CI perspective17:13
thorst_esberglu: lets do this least to start.17:14
thorst_lets turn on our CI for ceilometer proper17:14
thorst_and see what it does to our workload.17:14
fried_rice++ good call17:14
thorst_our CI is configured to install ceilometer now, right?17:14
thorst_you're sure?17:14
fried_riceWell, I used the CI's local.conf and got it when I stacked my victim, so...17:14
thorst_I mean it should be, I just don't recall...17:14
fried_riceenable_plugin ceilometer
fried_ricethat's it, right?17:15
thorst_well, but ceilometer-powervm too17:15
fried_riceenable_plugin ceilometer-powervm
thorst_yeah, I think that's all17:15
fried_ricewe're disabling ceilometer-aipmi17:15
fried_ricebut that's the only ceilometerish one we're disabling.17:16
thorst_yeah, I think that was the ipmi interface.17:16
esbergluthorst_: fried_rice: Regarding the official release bit. Since I already tagged releases we should be good for rc-3 right? Just adhere to that in the future17:16
thorst_which I don't think we want.17:16
fried_riceesberglu Yes, I think this is a Queens-and-forward statement.  Let me double check, though.17:16
esbergluthorst_: fried_rice: ceilometer has been running silently in CI for a while already17:21
esbergluSame with neutron17:21
thorst_:-)  so we're good bandwidth wise17:22
esbergluAre you wanting turn on voting like is done for nova proper?17:22
edmondswreading back17:22
fried_riceNot sure we're quite ready for that - I think we would want to initiate the discussion with gordc about integrating into ceilometer first.17:22
esberglufried_rice: Yeah I believe there are requirements before ceilometer would let us do that anyways. I can look into what those are17:23
thorst_fried_rice: we should chat with gordc at the PTG and get his views17:24
esbergluI'm assuming they are at least similar to nova17:24
thorst_or at least just have a quick chat in the ceilometer room.17:24
thorst_I don't think its super pressing17:24
fried_ricethorst_ Whenever you're ready.17:24
thorst_no, I meant at ptg17:24
fried_riceNo, but while it's on our minds.17:24
thorst_the physical room17:24
thorst_I know you're not used to those anymore17:24
fried_ricePlaces where I have to shave?17:25
thorst_as a software engineer, you never have to shave17:26
*** apearson has quit IRC17:29
edmondswok, caught up... sounds like the conclusion in openstack-meeting would be fine by us17:29
edmondswand we just need to figure out what to do for queens17:29
edmondswconversation with ceilometer at the PTG seems right to me17:29
thorst_yep.  Release thing and then a discussion with ceilometer at the PTG if we should fold in or not.  I'm 100% indifferent btw on that one17:30
edmondswlet me see if there's a ceilometer ptg etherpad we can throw this on...17:30
edmondswdoesn't seem to be one yet:
edmondswhmm... doesn't seem like there was one for Pike, either17:32
edmondswand now that I think of it, I don't remember seeing any of the ceilometer guys in Atlanta... I think they may have skipped it17:32
edmondswwe'd probably better initiate the convo on IRC, at least to ask whether they'll be at the PTG to discuss there17:33
edmondswwas there a reason we didn't just want to have the convo on irc en totale, thorst?17:33
fried_ricethorst_ probably wanted to make a slide deck about it.17:33
thorst_o man...too true too true.17:33
thorst_nah, it just didn't seem super pressing17:34
edmondswagree on that17:34
fried_riceYeah, exactly, I'm predicting if we don't do it now, we won't bother at the PTG, and it'll never happen.17:34
fried_riceWhich isn't necessarily a problem.17:34
fried_riceJust kind of might as well while we're thinking about it, kind of thing.17:34
fried_riceBut whatever.17:34
thorst_yeah, I guess my indifference is getting me on this.  I think they're indifferent too btw.  But I mean it can't hurt to ask.17:34
thorst_it's more like a verification that 'this is still cool, right?'17:35
fried_riceI think gordc is at lunch.  Let's revisit in an hour.17:35
edmondswlooks like their team meeting is Thursdays at 11am EDT17:35
edmondswopenstack-meeting channel17:36
thorst_yeah, we can poke in afterhis lunch and just give him a quick ask.  I'm basically booked for day in 1.5 hours though  :-(17:36
edmondswthink Friday? ;)17:37
thorst_meeting friday.17:37
thorst_but I'm going to go code for an the dream for a little bit17:37
*** esberglu has quit IRC18:11
*** openstackstatus has quit IRC18:11
*** openstack has joined #openstack-powervm18:13
*** esberglu has quit IRC18:15
fried_ricethorst_ gordc is back18:21
thorst_fried_rice: rgr18:24
*** apearson has joined #openstack-powervm18:29
*** esberglu has joined #openstack-powervm18:41
*** openstackstatus has quit IRC18:42
*** openstack has joined #openstack-powervm18:44
*** apearson has quit IRC19:08
*** smatzek has quit IRC20:30
*** apearson has joined #openstack-powervm20:42
*** thorst_ has quit IRC21:09
*** edmondsw has quit IRC21:23
*** edmondsw has joined #openstack-powervm21:24
*** edmondsw has quit IRC21:28
*** esberglu has quit IRC21:34
*** thorst has joined #openstack-powervm22:01
*** apearson has quit IRC22:01
*** thorst has quit IRC22:05
*** esberglu has joined #openstack-powervm22:07
*** esberglu has quit IRC22:11
*** thorst has joined #openstack-powervm22:31
*** fried_rice is now known as efried_zzz22:36

Generated by 2.15.3 by Marius Gedminas - find it at!