Wednesday, 2013-10-02

*** vipul is now known as vipul-away00:00
jeblairclarkb, jog0: maybe instead of e-r reading the zmq stream, we should have it be a gearman worker with a function that the logstash client can invoke.00:00
*** sarob has joined #openstack-infra00:01
jeblairclarkb, jog0: this occured to me since those two programs are basically implementing very similar logic to decide when to start working.00:01
clarkbyeah I have been considering the various options00:01
*** thingee is now known as thingee_zzz00:01
clarkbgearman seems nice because we can scale it if necessary00:02
*** emagana has joined #openstack-infra00:03
jeblairclarkb: and should be easy to make the logstash client call elastic-recheck when it's done stuffing things into logstash, right?00:03
*** emagana has quit IRC00:03
clarkbsort of. there is no indication of when elasticsearch is done indexing but we can ger it really close00:03
*** emagana has joined #openstack-infra00:04
clarkbwhich is better than the current poll situation today00:04
*** DennyZhang has quit IRC00:04
*** vipul-away is now known as vipul00:04
clarkbso ++ to gearman00:05
*** Cyril2 has joined #openstack-infra00:06
*** sarob has quit IRC00:07
*** jpipes has quit IRC00:08
*** sarob has joined #openstack-infra00:08
*** sarob has quit IRC00:10
*** sarob has joined #openstack-infra00:10
*** AlexF has quit IRC00:15
jog0++ from me too00:18
jeblairjog0: is there a script to run the queries and output counts?00:20
jog0jeblair: there will be in 10 minutes00:21
jog0jeblair: thats on the TODO list00:21
jog0will do it now00:21
*** UtahDave has joined #openstack-infra00:21
*** svarnau has quit IRC00:25
*** matsuhashi has joined #openstack-infra00:25
jog0clarkb: ElasticSearch is still slow00:34
jog0and elastic-recheck is hung again00:34
*** emagana_ has joined #openstack-infra00:39
clarkbjog0 ya until my longer timeout patch goes in and your change to timeout goes in this will be a problem00:41
jog0clarkb: ahh00:41
*** julim has quit IRC00:42
jog0and your patch hasn't landed yet?00:42
jog0and the load isn't down?00:42
*** emagana has quit IRC00:43
clarkbmy patch hasn't even been reviewed I don't think00:43
clarkbI just pushed it :)00:43
clarkband load may be related to the crazy find we run00:44
jeblairclarkb: i reviewed it!00:44
jeblairclarkb: what load?00:44
jeblairclarkb: and we don't run it during the week00:45
clarkbjeblair hypothesis is that load domewhere between slave and static makes file copy slow00:45
jeblairclarkb: did i miss a conversation?00:45
clarkbwhat is happening is after 60 seconds console.html is not copied to static. sort of I mention it briefly in my commit message I think00:46
jeblairclarkb: possibly the jenkins master, since all file copies are funneled through it.  and we do hundreds of them at once.00:46
clarkbbut I wanted more time to investigate before digging deeply00:46
clarkbjeblair oh good point00:46
*** dcramer_ has joined #openstack-infra00:48
*** nosnos has joined #openstack-infra00:49
jog0jeblair: turns out https://review.openstack.org/#/c/49246/100:51
jog0generates what you want, running it now00:51
jeblairjog0: cool00:52
jog0these queries are a little slow when open ended00:54
*** ryanpetrello has joined #openstack-infra00:54
jgriffithclarkb: jeblair quick question... was going to propose on ML but thought I'd ask in here first :)00:56
jgriffithseems we've identified a number of bug fixes that are actuallyin progress to fix some of the failures that are choking up gates00:56
jgriffithany chance we could/should start systematically ONLY allowing those changes in for the time being?00:57
*** adalbas has quit IRC00:58
*** darraghb has joined #openstack-infra00:58
*** weshay has quit IRC00:59
jog0jgriffith: ++01:00
jog0ttx: kicked that idea around before too01:00
jog0ttx kicked *01:01
jgriffithahh...ok, so not totally off the wall then01:01
jog0jeblair jgriffith: here are the top offending bugs http://paste.openstack.org/show/47807/01:01
jog0unsorted01:01
jgriffithjog0: here's my homework assignment http://paste.openstack.org/show/47798/01:02
jgriffithjog0: trying to get the last cinder one in since 9 this morning :(01:02
jgriffithbut not sure how frequent it is of late01:02
jog0nice01:02
*** schwicht has quit IRC01:03
jgriffithjog0: BTW I think those top list items have changed through the day01:04
jgriffithanyway...01:04
jog0jgriffith: I just ran the check momemnts ago01:05
jog0so the paste I gave is fresh01:05
jgriffithjog0: oh... then never mind :)01:05
clarkbjog0 jgriffith I think you fix that socially... we could extend zuul to have change based acls but asking folks to play nice seems simple01:05
*** melwitt has quit IRC01:05
jgriffithclarkb: seems like it, however i"m finding otherwise01:05
jgriffithie "My driver has a misspelled commet, here's my change"01:06
jog0jgriffith:  bug 1226337 at 38901:06
uvirtbotLaunchpad bug 1226337 in cinder "tempest.scenario.test_volume_boot_pattern.TestVolumeBootPattern flake failure" [Undecided,Fix committed] https://launchpad.net/bugs/122633701:06
jog0bug 1229475 at 500!01:06
uvirtbotLaunchpad bug 1229475 in python-neutronclient "terminate_instance(): RuntimeError: Second simultaneous read on fileno 16 detected" [Undecided,In progress] https://launchpad.net/bugs/122947501:06
jog0bug 1223469 at 10901:07
uvirtbotLaunchpad bug 1223469 in cinder "Volume create tgtadmin fails with error code 22" [High,Fix committed] https://launchpad.net/bugs/122346901:07
jgriffithjog0: got ya01:08
jgriffithjog0: so the good thing is 1223469 shouldn't be causing failures anymore01:08
jgriffithjog0: but at this point it's small taters01:08
*** ryanpetrello has quit IRC01:08
jog0impressively 2 should be closed and one is on its way01:08
jgriffithcool... I just be patient :)01:09
jgriffithI really want to get Cinder off of your list ;)01:09
jog0so after 2 weeks when we remove the logs from logstash we will see the bug hasn't happend and remove it from elastic-recheck01:09
jog0jgriffith: ^_^01:09
darraghbhas anyone had a chance to see if the single entry point assumption in the following change for JJB https://review.openstack.org/#/c/49102/ is valid?01:09
jog0I know there are some nova issues not on that list01:10
jog0and maybe neutron01:10
jgriffithjog0: so maybe I need to change the code to not print an error message if the recovery works01:10
jog0that makes sense01:11
*** DennyZhang has joined #openstack-infra01:12
clarkbdarraghb: I haven't had a chance to look01:12
jog0jgriffith: ther is also this one bug 120534401:13
uvirtbotLaunchpad bug 1205344 in nova "mkfs error in test_stamp_pattern" [Medium,Confirmed] https://launchpad.net/bugs/120534401:13
jog0which we are disabling the test for01:13
jgriffithjog0: got ya01:13
jog0and will re-enable early in Icehouse01:13
jog0that maybe a nova issue too?01:14
jgriffithYeah, I was looking at that one, it's a big funky01:14
*** changbl has joined #openstack-infra01:14
jgriffithjog0: not a ton to go on01:15
jog0jgriffith: :( it keeps happening though01:15
jog0but this is post havana01:15
jgriffithsighh01:16
*** ryanpetrello has joined #openstack-infra01:17
jog0jgriffith: anyway I have a plane to catch01:18
jog0just think when the gate is stable and we do a better job of catching these issues you won't have this pain again01:19
jgriffithjog0: safe travels01:19
*** ryanpetrello has quit IRC01:20
*** ryanpetrello has joined #openstack-infra01:22
mriedemdoes anyone know why this doesn't fall back to get thrift from pypi.openstack.org? http://logs.openstack.org/39/48739/4/check/gate-requirements-install/6a6ce59/console.html01:28
*** sarob has quit IRC01:28
mriedempypi.python.org is not responding, but the thrift package is in the openstack mirror01:28
*** sarob has joined #openstack-infra01:29
clarkbmriedm: requirements doesn't use our mirror01:32
clarkbits a chicken and egg problem01:32
*** sarob has quit IRC01:32
clarkbrequirements builds the mirror so needs to be tested against pypi01:32
dstufft I see bumping the timeout just made it 50301:33
dstufftI was worried it might01:33
dstufftI'm sorry for that ^, upstream Fastly is still investigating01:33
*** dims has quit IRC01:40
*** emagana_ has quit IRC01:52
*** ArxCruz has quit IRC01:52
*** sandywalsh has joined #openstack-infra01:53
hub_capmordred: its coming, albeit rather slowly https://gist.github.com/hub-cap/678788501:54
*** sarob has joined #openstack-infra01:55
*** ArxCruz has joined #openstack-infra01:56
*** ryanpetrello has quit IRC01:58
*** DennyZhang has quit IRC02:10
*** sarob_ has joined #openstack-infra02:15
*** jerryz has quit IRC02:17
*** sarob has quit IRC02:18
*** sarob_ has quit IRC02:20
*** ryanpetrello has joined #openstack-infra02:25
mriedemclarkb: ok, thanks, will recheck when i can get the tar myself02:26
darraghbHoping to resurrect the thread patch I uploaded previously that was abandoned https://review.openstack.org/#/c/45371/102:33
darraghbCould do with some suggestions on how to ensure the CI tests will use multiple threads to catch issues while the code defaults to single threaded behaviour to ensure end use of the change requires explicit enabling?02:34
*** AlexF has joined #openstack-infra02:36
*** dkehn_ has joined #openstack-infra02:37
*** dkehn has quit IRC02:41
*** markmcclain has joined #openstack-infra02:42
*** gyee has quit IRC02:52
anteayaI'm looking for the reason for this grenade-devstack-vm failure and so far I can't find it: https://review.openstack.org/#/c/48971/02:54
anteayaanyone willing to take a peek and offer an opinion?02:54
*** AlexF has quit IRC02:59
mriedemanteaya: maybe because it can't find the test results log files it's looking for?03:03
*** ryanpetrello has quit IRC03:03
mriedem[SCP] '**/subunit_log.txt.gz' doesn't match anything: '**' exists but not '**/subunit_log.txt.gz'03:03
anteayareally?03:03
anteayathat would create a failure?03:04
clarkbno that won't cause a failure03:04
anteayaI need to get it back in the queue, but somehow reverify no bug would feel like I was missing something03:06
anteayabut for the life of me, I can't find what caused that test to fail03:06
mriedemanteaya: clarkb: i'm just comparing to this which didn't fail: http://logs.openstack.org/71/48971/2/gate/gate-grenade-devstack-vm/632df89/03:07
mriedemthat was from a previous run, it doesn't fail and the log gz files are there03:07
mriedemalthough it has a ton more logs here too: http://logs.openstack.org/71/48971/2/gate/gate-grenade-devstack-vm/632df89/logs/03:07
anteayayes03:08
clarkbgrenade is returning 1 for some reason03:08
anteayaa large description in the amount and format of the logs03:08
anteaya2013-10-02 02:43:13.040 | + return 1 in console.html03:09
anteayaa bit slim to go on for debugging03:09
anteayaat this timestamp as well: 2013-10-02 02:43:13.027 | + return 103:10
mriedemanteaya: i see 138 hits for "return 1" :)03:10
anteaya2013-10-02 02:43:13.015 | + return 0 above that03:10
anteayathat is quite a few03:11
anteayamriedem: how many do you get in this file: http://logs.openstack.org/71/48971/2/gate/gate-grenade-devstack-vm/632df89/console.html03:12
anteayaI get about 19003:13
anteayaso return 1 might not be a good indicator of what is unusual in the console.html for the failure03:13
mriedem191 for me03:14
mriedem~same03:14
mriedemoff topic, but this could use another infra +2, i've been rechecking the hell out of it today: https://review.openstack.org/#/c/48739/03:14
mriedemclarkb: ^?03:14
anteayaI'm going to swing 48971 back onto the pile with no bug03:15
anteayaI'll hang my head in shame as I do it03:15
mriedemi think it was already rechecked03:16
mriedemreverified i mean03:16
anteayaoh morganfainberg already did it03:16
anteayayup03:16
anteayaI'll hang my head in shame another time03:16
mriedemanteaya: yeah, about 20 min ago03:16
anteayafine03:16
anteayaI couldn't find the failure anyhow03:16
mriedemanteaya: let that be your shame03:17
anteayaI'm not very experienced at reading the grenade failure files03:17
anteayayay, I have a target for the shame03:17
anteayaso someday I will learn how to read the grenade failure files03:17
clarkbmriedem: I would look but am on a phone. about to afk03:18
mriedemclarkb: ok, thanks anyway03:19
*** hogepodge has quit IRC03:26
lifelessclarkb: how do I go about getting a service account ?03:26
*** mriedem has quit IRC03:28
clarkblifeless: file a bug/request one via email/irc and provide pubkey03:30
jog0jgriffith: are you still thinking of asking the ML to be careful about approving patches?03:30
jog0If so I am all for it03:31
lifelessclarkb: so I'd like one03:31
*** UtahDave has quit IRC03:32
lifelessclarkb: 'tripleo-cd-bot', pubkey is linked from https://etherpad.openstack.org/tripleo-mvp103:32
clarkblifeless: ok. I would help now but am trying to get off this phone :) probably easiest since most of us are afk to submit a bug or send email to infra list03:32
*** darraghb has quit IRC03:33
lifelessmaybe mordred will do it when he gets back :)03:33
lifelessI will file bug now03:33
jgriffithjog0: well... not sure but it looks like we may be rounding a corner03:34
jgriffithunless I'm just being overly optimistic03:34
lifelessclarkb: https://bugs.launchpad.net/openstack-ci/+bug/123396103:35
uvirtbotLaunchpad bug 1233961 in openstack-ci "Please add a service account" [Undecided,New]03:35
clarkblifeless: perfect thanks03:36
mordredwhat did I miss?03:42
*** Ryan_Lane has joined #openstack-infra03:44
jgriffithjog0: last four Cinder patches have passed, haven't seen that in a while03:49
jgriffithjog0: like I said, don't want to jinx it, but maybe we're rounding the corner03:49
*** darraghb has joined #openstack-infra03:49
*** dkliban has joined #openstack-infra03:50
lifelessmordred: https://bugs.launchpad.net/openstack-ci/+bug/123396103:50
uvirtbotLaunchpad bug 1233961 in openstack-ci "Please add a service account" [Undecided,New]03:50
*** Ryan_Lane has quit IRC03:50
mordredjgriffith: woot03:54
mordredlifeless: nod03:54
anteayajgriffith: oh I hope so03:54
anteayaI need to hear that somebody is happy today03:54
* jgriffith is always happy03:56
jgriffithmordred: hey.. btw, how do I recheck smokestack (only had to ever do it once and that was like a year ago)03:56
anteayayay03:57
mordredjgriffith: there's no way to without poking dan03:58
mordredjgriffith: we don't run smokestack03:58
jgriffithmordred: that's whatI thought.. is it non-voting?03:58
mordredyes03:58
jgriffithmordred: excellent.. thanks03:59
morganfainberganteaya, the granade error was actually in the build, wasn't anything really wrong with the code (mysql timed out it looked like)03:59
anteayamorganfainberg: can you show me the line/timestamp where you saw that?03:59
morganfainbergsure, sec.03:59
morganfainberganteaya, 2013-10-02 02:22:59.965 in console.html also notice how the other logs were really really empty compared to others.04:00
morganfainbergother grenade runs that is04:00
morganfainbergoh hah, wasn't even mysql time out was a total 60 minute timeout04:01
anteayayes I did notice the absence of logs and content04:01
morganfainberghaven't seen that much in granade tbh04:01
anteayame either04:01
anteayaa 60 minute timeout?04:01
morganfainbergyeah, thats a long time to be spinning on the build.04:02
anteayaI am about to stick my foot in my mouth, the build of what04:02
*** UtahDave has joined #openstack-infra04:02
anteayawhat was building04:02
anteayathat was given 60 minutes to do so and couldn't manage it04:02
anteayadevstack?04:03
morganfainberganteaya, setup maybe?04:03
morganfainberganteaya, thats what it looked like04:03
morganfainberganteaya, it looked transient at worst, just a bad setup.04:03
morganfainberganteaya, hence why i went with no bug.  (I haven't seen that happen really ever)04:03
anteayawell it got it back on the pile04:03
anteayaso that is good04:03
anteayabut I find this ... odd04:04
*** tkammer has joined #openstack-infra04:04
morganfainberganteaya, yep. i'll babysit it some more tonight.  I've been babysitting RC bugs through gate for keystone for the last couple weeks at night04:04
anteayanow it might be odd because I don't know how to read grenade logs04:04
anteayacool04:04
morganfainberganteaya, in this case, the lack of logs make it a particularly bad example :P04:04
anteayaI'm also babysitting the glance one04:04
anteayathat is true04:04
anteayaso we are on the night watch04:05
morganfainberganteaya, otherwise granade isn't a whole lot different than most tempest like logs (devstack)04:05
morganfainberganteaya, Winter is coming?04:05
morganfainberg>.>04:05
anteayaapparently04:05
anteayamy nephew has a shirt that says so04:05
morganfainberganteaya, but i didn't want to take the black :(  i don't want to go to the wall :P04:05
* morganfainberg stops w/ the ASoIF (or GoT) references...04:05
anteayaI feel I am missing something04:06
anteayaI feel my nephew would get it though04:06
morganfainberganteaya, Game of Thrones04:06
anteayaah04:06
anteayano idea04:06
anteayaI can tell you about guitarzan though04:06
morganfainberganteaya, if you like low-fantasy and very political (kings, queens, subterfuge, royalty, etc) its a decent series to read.  And if you like HBO and the "HBO-Style" shows, it's and entertaining watch.04:07
morganfainbergbut the show/books aren't for everyone04:08
anteayayay 49029 succeeded04:08
* morganfainberg cheers for 4902904:08
anteayago swift and horizon patches04:08
*** harlowja has quit IRC04:08
anteayamerge, merge, merge04:08
morganfainberghehe04:08
anteayait has been so long since I watched a TV series04:08
anteayaI get disappointed in the writing04:09
morganfainbergat least we aren't up to the oslo-policy sync reverify's for the keystone merges yet. https://review.openstack.org/#/c/48111/ (it got silly)04:09
morganfainberganteaya, ah yeah. i don't have TV service anymore.  I only watch the shows I want (~2) via like itunes.04:09
anteayawoo Sept 28 was a memorable time for you watching ZuulTV04:10
anteayaI don't blame you04:11
morganfainberghehe04:11
anteayaactually the writing for ZuulTV is quite good04:11
anteayaproduction quality is impressive04:11
morganfainberganteaya, hehehehe ++04:11
anteaya:D04:11
*** markmcclain1 has joined #openstack-infra04:15
*** markmcclain has quit IRC04:15
*** markmcclain1 has quit IRC04:19
*** Ryan_Lane has joined #openstack-infra04:19
*** markmcclain has joined #openstack-infra04:19
*** emagana has joined #openstack-infra04:20
*** Ryan_Lane has quit IRC04:23
*** dkliban has quit IRC04:23
*** tkammer has quit IRC04:28
*** tkammer has joined #openstack-infra04:29
*** esker has joined #openstack-infra04:41
lifelessmordred: I am really wondering how hard it would be to tell jjb to make three jobs for each gate job definition04:47
lifelessmordred: and zuul to schedule all three04:47
lifelessas a low-overhead variant of 'run N', just to see what it scares up04:48
mordredlifeless: so, I believe we've landed the patch now to allow us to tell zuul to schedule multiple runs of a given job for a patch04:49
mordredlifeless: before, we were explicitly collapsing duplicates in the job list04:49
*** Ryan_Lane has joined #openstack-infra04:49
morganfainbergmordred, running multiples could be beneficial04:49
mordredmorganfainberg: yes. it's definitely on the list04:49
mordredonce we get the current thing under control, we'd liek to put multiple runs in to try to prevent new things from coming in04:50
morganfainbergmordred, i'm going to start taking a hard look at some of the infra stuff once i have a few cycles, see if i can contribute some help along those lines.04:50
morganfainbergor something in here :)04:50
mordredmorganfainberg: yay!04:50
mordredmorganfainberg: I like it when people contribute!04:51
morganfainbergmordred, yah! and i like contributing!04:51
*** Ryan_Lane has quit IRC04:54
*** SergeyLukjanov has joined #openstack-infra04:58
*** boris-42 has joined #openstack-infra04:59
markmcclainis there  something wrong with jenkins01? it has been SCP'ing horizon doc files for nearly 90mins05:01
morganfainbergmarkmcclain, it's also been slow loading the consoles/info pages on the jobs05:02
*** UtahDave has quit IRC05:03
anteayait seems to be processing the files05:09
anteayaperhaps it just has a lot of files to process05:09
markmcclainno more than files than the other horizon doc jobs that have already completed05:13
morganfainbergmarkmcclain, maybe just a bad vm/slave that is misbehaving from an I/O standpoint?05:14
mordredmarkmcclain: quantumclient released05:14
markmcclainmordred: thanks05:14
anteayamarkmcclain: true it may be slow compared to other vms, but it is still processing05:15
anteayaI'm not a fan of waiting either, but it isn't stuck05:16
morganfainberganteaya, but uploading 3-4 files per minute is … very slow.05:16
anteayayes05:16
morganfainbergjust as long as it completes and doesn't fail.05:16
anteayaright05:16
morganfainbergi'll be ok with it05:16
anteayasince there is no intervention that will make it go faster05:16
anteayaof which I am aware05:16
morganfainbergI shall glare menacingly05:17
morganfainbergmaybe it'll go faster05:17
anteayaglare away05:18
anteayaI hope it works05:18
anteayaI'm browsing reddit05:18
anteayalet's see if that helps05:18
anteayamy ability to think passed long ago05:18
morganfainberganteaya, i'm getting my grumpycat fix in :P05:19
anteayanot a meme I ever felt akin too05:19
anteayabut if grumpycat does it for you, knock yourself out05:20
*** Ryan_Lane has joined #openstack-infra05:20
morganfainberganteaya, not the meme , but the actual cat videos pics w/o the text :)05:20
anteayathis one I don't know05:20
*** Ryan_Lane has quit IRC05:24
*** tvb|afk has joined #openstack-infra05:25
*** tvb|afk has quit IRC05:30
anteayamordred, it is like you just tailed off and hit send on that one05:42
*** jaypipes has joined #openstack-infra05:46
anteayahow can this be merged: https://review.openstack.org/#/c/49006/05:46
anteayawhen it is at the end of the gate queue and failing spectacularly?05:46
anteayamordred ^05:46
*** pblaho has joined #openstack-infra05:47
SergeyLukjanovanteaya, Patch Set 1: Verified; Looks good to me (core reviewer); Approved05:47
anteayastatus is merged05:48
anteayaUploadedSep 30, 2013 12:42 PM05:48
anteayaUpdatedOct 2, 2013 1:13 AM05:48
anteayaStatusMerged05:48
anteayabut 1:13 was when it was approved05:48
anteayaand it is in the bottom of the gate queue05:49
anteayafailing on every devstack test so far05:49
mordredanteaya: I force merged it05:49
mordredanteaya: it will never pass tests05:49
anteayaoh okay05:49
anteayaI see05:50
anteayalike it isn't passing tests now05:50
mordredit's the quantumclient branch, which will be spectacularly broken in our rig - but we don't care :)05:50
mordredyah05:50
anteayaoh we don't care05:50
anteayatomorrow I will ask why we don't care05:50
anteayabut not tonight05:50
*** Ryan_Lane has joined #openstack-infra05:50
*** odyssey4me has joined #openstack-infra05:50
anteayatonight I want the squirrels on the Jenkins01 doc scp job to keep rolling05:51
mordredmmm. squirrels05:52
*** tvb|afk has joined #openstack-infra05:54
*** Ryan_Lane has quit IRC05:54
*** SergeyLukjanov has quit IRC05:55
anteayaoh this glacial doc job had better pass05:57
jog0anteaya: I am watching it too05:57
anteayawe will hopefully clear about about 19 patches05:57
anteayajog0: any thoughts on why that doc job is taking so long?05:57
jog0last I saw it was on scp for the past hour or so but not sure05:58
anteayayup05:58
anteayastill scp'ing05:58
jog0lol05:58
jog0wonder if its rax issue?05:58
*** esker has quit IRC05:58
jog0other pieces are slow today too05:58
anteayait is a rax vm?05:58
jog0anteaya: not sure05:58
anteayaI havent' been paying attention to speed today05:59
*** odyssey4me has quit IRC05:59
*** esker has joined #openstack-infra05:59
anteayafor the logs we are discussing this patch: https://review.openstack.org/#/c/44724/05:59
anteayaall done 'cept gate-horizon-docs06:00
anteayawhich is holding up everything else06:00
anteayathough still running06:00
jog0anteaya: pretty amazing its still running06:00
anteayait is06:00
anteayathe 20 patches behind it are all finished (passed or failed)06:01
Alex_Gaynorit's got like 20 minutes until it times out I believe06:04
morganfainbergAlex_Gaynor, i hope it doesn't time out...06:06
*** odyssey4me has joined #openstack-infra06:08
anteayaAlex_Gaynor: how long does it have total?06:09
Alex_Gaynoranteaya: I think the timeout is 3hours?06:09
anteayawow06:09
anteayayea, we are close to that06:10
morganfainbergis there anyway we can prevent the timeout from happening?06:10
anteayasince that is a non failing item I'm betting that it will be a gate reset06:10
anteayaI have no power06:10
anteayait is in the code I do believe06:11
anteayanot a trigger on the vm06:11
morganfainbergnod06:11
anteayaI still feel awful sitting here, watching the timeout about to happen06:11
openstackgerritAlex Gaynor proposed a change to openstack-infra/config: Handle null in this function  https://review.openstack.org/4931206:12
anteayaI'm curious as to the reason why for 4931206:14
anteayasince we are here06:14
Alex_Gaynoranteaya: ugh, I should improve that commit message06:15
anteayak06:15
openstackgerritAlex Gaynor proposed a change to openstack-infra/config: Handle null in this function  https://review.openstack.org/4931206:16
*** tvb|afk_ has joined #openstack-infra06:16
anteayamordred you still up?06:19
*** tvb|afk has quit IRC06:20
mordredanteaya: yup06:20
anteayagrand06:20
*** Ryan_Lane has joined #openstack-infra06:20
*** shardy_afk is now known as shardy06:23
*** odyssey4me2 has joined #openstack-infra06:25
*** odyssey4me has quit IRC06:28
anteayaBuild has been executing for 3 hr 4 min on precise706:28
anteayaand still running06:28
anteayawhew06:28
*** Ryan_Lane has quit IRC06:29
morganfainbergvery close to being done06:29
anteayaoh i hope we make it06:29
*** odyssey4me2 has quit IRC06:29
Alex_GaynorApparently it's not a 3 hour cutoff?06:30
morganfainberg~60 more files to upload06:30
*** tvb|afk_ is now known as tvb06:31
morganfainbergAlex_Gaynor i think that is because it's in upload mode06:31
*** tvb has quit IRC06:31
*** tvb has joined #openstack-infra06:31
morganfainbergdifferent phase of the build06:31
Alex_Gaynormorganfainberg: oh, can it not time out at this stage?06:31
morganfainbergnot "execution" of the build06:31
Alex_GaynorI guess that makes sense, it's outside the main script06:31
morganfainbergAlex_Gaynor, no idea, but jenkins handles that separately06:31
morganfainbergso that is my guess06:31
morganfainbergAlex_Gaynor, in either case this is a bit nerve wracking because of how many successful gate jobs are behind it06:33
Alex_Gaynorindeed06:33
anteaya20 completed patches behind it06:35
anteaya16 successful06:35
morganfainbergwoo, last 8 files or so06:38
anteayago go go06:39
anteayayayaya06:39
morganfainbergand success06:39
anteayafinished success06:39
anteayayay06:39
anteayathere goes horizon06:40
morganfainbergand phew, that means the RC1 patch made it in for keystone :)06:40
*** emagana has quit IRC06:40
anteayayay06:41
sldwhat patch?06:41
morganfainberguhoh06:41
anteaya4897106:41
morganfainbergis zuul … unhappy?06:41
anteayazuul just needs a moment06:41
anteayarecalculating06:42
morganfainbergok06:42
anteayaand uploading logs to comments06:42
anteayagive him a minute06:42
anteayahe has had to entertain us for over 3 hours06:42
anteayathe god need s a break06:42
sldthis?  https://review.openstack.org/#/c/48971/  ?06:43
morganfainbergsld yeah06:43
anteayagate-horizon-docs SUCCESS in 3h 16m 20s06:46
morganfainbergwoot06:47
anteayaI'm so glad we made it06:47
AJaegercongrats!06:47
anteayaAJaeger: thanks06:47
anteayafeels like a victory06:47
anteayathough I did nothing but watch zuulTV06:48
anteayaand click06:48
sld3 hours?  geez.06:48
anteayayeah, no kidding06:48
anteayacompared with this horizon patch: https://review.openstack.org/#/c/49249/06:48
anteayagate-horizon-docs SUCCESS in 35m 32s06:49
AJaegeranteaya, it's not only the zuulTV but all the buildup over the last 6 months. But without zuulTV you would had less fun ;)06:50
anteayamuch less fun indeed06:50
anteayayes, the culmination of a grand effort06:50
*** darraghb has left #openstack-infra06:51
*** afazekas_no_irq has joined #openstack-infra06:55
*** Ryan_Lane has joined #openstack-infra06:55
*** uvirtbot has quit IRC06:56
*** odyssey4me has joined #openstack-infra06:56
*** uvirtbot has joined #openstack-infra06:56
*** Ryan_Lane has quit IRC06:59
*** odyssey4me2 has joined #openstack-infra07:00
*** vipul is now known as vipul-away07:02
*** odyssey4me has quit IRC07:02
*** boris-42 has quit IRC07:06
jd__zuul doesn't seem very active according to http://status.openstack.org/zuul/07:11
jd__I hope it's just a glitch on the web interface :)07:11
anteayawe just waited a long time for a job at the head of the queue to finish07:12
morganfainbergjd__, i am hoping it's just catching up from the ~16 patchest that completed gate all at once.07:12
anteayaI am hoping that the pause is due to writing logs and recalculating the pipeline07:13
jd__anteaya: yeah but now… I've approved a patch and asked for a recheck 5 minutes ago already07:13
anteayayeah what morganfainberg said07:13
jd__i'm just scared nothing shows up, not even the event queue length07:13
anteayaoh goodness07:14
anteayaI just refreshed my page07:14
* ttx yawns07:14
morganfainberganteaya, yeah07:14
anteayano queue07:14
morganfainbergthat is what made me go "uhhhhhh"07:14
anteayamorning ttx07:14
lifelessit'll be ttx again07:14
ttxanteaya: still up?07:14
lifelesshe broke it yesterday when he got up :)07:14
anteayayeah07:14
anteayait be broke07:15
morganfainberg=/07:15
lifelesscheck the json07:15
anteayahttp://zuul.openstack.org/status.json07:15
anteayasaved a copy07:16
anteayanot sure what to do with it though07:16
ttxanteaya: was the same yesterday07:16
anteaya:/07:16
anteayathis was the queue07:16
anteaya44724,4 running07:16
anteaya49249,107:16
anteaya49155,107:16
anteaya47969,807:16
anteaya48608,107:16
lifelesswell, it's not empty :)07:16
anteaya47463,307:16
anteaya49029,1 glance passed07:16
anteaya46747,507:16
anteaya47503,1607:16
anteaya49266,107:16
anteaya49083,107:17
anteaya48794,207:17
anteaya49105,107:17
ttxI saved json for jeblair but the json wasn't really weird07:17
anteaya48968,1 failed07:17
anteaya48628,7 failed07:17
anteaya49275,107:17
anteaya49019,307:17
anteaya49191,1 failed07:17
anteaya48993,207:17
anteaya48971,2 keystone passed07:17
anteaya49287,1 failed07:17
anteaya47453,2 running07:17
anteaya48836,107:17
anteaya49233,107:17
lifelessanteaya: uhm enough :)07:17
anteaya49058,307:17
anteaya48891,107:17
lifelessanteaya: pastebin is better for that07:17
anteaya47130,1107:17
anteaya49088,307:17
anteaya48387,507:17
anteaya49006,1 failed07:17
anteaya48528,807:17
ttxanteaya: i'm reading from the json :)07:17
anteayaI think perhaps there were so many failed branches zuul might have given up07:17
anteayalifeless: yes you are correct07:17
anteayaI'm tired and lazy, my bad07:18
lifelessanteaya: zuul creates that json; the failure is only in the UI07:18
anteayahttp://paste.openstack.org/show/47821/07:18
anteayaI guess what I am saying is that patch 14 and 15 failed07:19
*** SergeyLukjanov has joined #openstack-infra07:19
anteayaand patch 1807:19
anteayaand patch 2107:19
anteayaso 3 different branches for nnfi07:19
anteayaI wonder if that was too much for it to calculate07:19
anteayaso when 44724 completed, it took a long break07:20
*** flaper87|afk is now known as flaper8707:20
anteayaeverything prior to patch 22 had completed07:20
ttxanteaya: go to bed, we'll babysit it from here :)07:20
anteayak07:20
anteayathanks good night07:20
ttxgood night!07:20
anteaya:D07:21
*** anteaya has quit IRC07:21
*** davidhadas has quit IRC07:23
ttxjd__: zuul status page was broken in the same way same time yesterday07:26
*** Ryan_Lane has joined #openstack-infra07:26
*** AlexF has joined #openstack-infra07:29
*** Ryan_Lane has quit IRC07:30
*** vipul-away is now known as vipul07:30
lifelessttx: did you file a bug yesterday ?07:31
*** odyssey4me2 has quit IRC07:32
*** jcoufal has joined #openstack-infra07:34
*** DinaBelova has joined #openstack-infra07:41
ttxlifeless: no, I only informally signalled it to jeblair07:42
ttxand sent him a couple status.jsons07:42
ttxlifeless: I wonder if it's not time-sensitive. Zuul's display waits for jeblair to go to bed and then fails.07:43
Mithrandirso we should just get jeblair a coffee drip?07:43
*** flaper87 is now known as flaper87|afk07:43
ttxMithrandir: that's a known solution to many problems.07:44
ttxMithrandir: i've been encouraging jeblair to move to France.07:44
ttxthat would solve it.07:44
Mithrandirbecause of the stronger coffee?07:45
ttxMithrandir: stronger coffee and the resulting 24-h core infra coverage07:45
*** odyssey4me has joined #openstack-infra07:47
lifelessttx: ITYM infra-admins or whatever the root team name is :)(07:47
ttxlifeless: starting to wonder if it's actually not the periodic jobs f*cking things up. They finish around 0600 UTC which is fail time07:48
ttxat the very least they add complexity to the display07:48
*** yassine has joined #openstack-infra07:48
lifelessit is rather coincidental07:49
ttxlifeless: because the rest of the gate is not significantly busier now07:49
openstackgerritlifeless proposed a change to openstack-dev/cookiecutter: Ignore editor files.  https://review.openstack.org/4931807:50
*** Ryan_Lane has joined #openstack-infra07:56
*** flaper87|afk is now known as flaper8707:59
*** jcoufal_ has joined #openstack-infra07:59
morganfainbergugh i think something went wonky w/ zuul's gate queue.  prior to the UI going out, https://review.openstack.org/#/c/48971/ and it's preceeding list appeared to have hit full success. now the json is claiming it's running again08:00
*** Ryan_Lane has quit IRC08:01
morganfainbergeither ui was misleading or something partially reset the queue somewhere between the 3hr horizon doc job and ~16 successes down the list08:01
*** hashar has joined #openstack-infra08:01
*** jpich has joined #openstack-infra08:01
*** jcoufal has quit IRC08:02
*** davidhadas has joined #openstack-infra08:03
*** thomasbiege has joined #openstack-infra08:07
*** thomasbiege has quit IRC08:09
*** flaper87 is now known as flaper87|afk08:09
*** flaper87|afk is now known as flaper8708:12
*** davidhadas has quit IRC08:13
*** boris-42 has joined #openstack-infra08:15
*** derekh has joined #openstack-infra08:16
*** AlexF has quit IRC08:17
*** davidhadas has joined #openstack-infra08:19
*** basha has joined #openstack-infra08:23
*** resker has joined #openstack-infra08:26
*** tvb has quit IRC08:26
*** Ryan_Lane has joined #openstack-infra08:27
*** matsuhashi has quit IRC08:28
*** flaper87 is now known as flaper87|afk08:28
*** matsuhashi has joined #openstack-infra08:28
*** esker has quit IRC08:29
*** Ryan_Lane has quit IRC08:31
*** OlivierSanchez has joined #openstack-infra08:32
*** AlexF has joined #openstack-infra08:32
*** mrmartin has joined #openstack-infra08:32
*** sld has quit IRC08:33
*** dizquierdo has joined #openstack-infra08:36
openstackgerritlifeless proposed a change to openstack-infra/gerritlib: Simplify GerritWatcher creation.  https://review.openstack.org/4932608:37
openstackgerritlifeless proposed a change to openstack-infra/gerritlib: Ignore editor files.  https://review.openstack.org/4932708:37
*** bnemec has quit IRC08:39
*** flaper87|afk is now known as flaper8708:39
*** johnthetubaguy has joined #openstack-infra08:41
*** AlexF has quit IRC08:42
*** DinaBelova has quit IRC08:44
hasharlifeless: awesome :-)08:44
hasharlifeless: I was wondering yesterday wether gerritlib is still used/maintained08:45
*** basha has quit IRC08:46
ttxzuul status is back08:47
lifelesshashar: well, I have no idea :)08:47
*** davidhadas_ has joined #openstack-infra08:47
lifelesshashar: but zuul does use it08:48
*** johnthetubaguy1 has joined #openstack-infra08:48
hasharlifeless: gasp, yet another dependency I will have to package :D08:48
lifelesshashar: package?08:49
*** yolanda has joined #openstack-infra08:49
*** johnthetubaguy has quit IRC08:50
*** davidhadas has quit IRC08:50
yolandahi, i'm trying to setup permissions on non-interactive users group using git, but i find that non-interactive users group is not listed in groups file. As a consequence, permissions aren't updated until i save them from the UI, then Non-Interactive Users group is created. Any idea on what can be happening? other groups such as Administrators, Anonymous, etc... are in groups file08:51
jd__ttx: that's good news :)08:51
jd__and I see a lot of success08:51
jd__it's like the sky went cleared out of the (bad) clouds08:52
hasharlifeless: at Wikimedia we rely almost exclusively on Debian packages.  So whenever there is new python module, I got to package it to be able to install the module in production :-]08:55
ttxjd__: a few things are running extremely slowly though. That test we are waiting on at the top of the queue is not even finished installing devstack08:56
ttxbeen running for more than one hour now08:56
jd__:-/08:57
jd__ceilometer should be ready soon otherwise08:57
*** Ryan_Lane has joined #openstack-infra08:57
lifelesshashar: ah08:58
*** DinaBelova has joined #openstack-infra08:59
*** davidhadas has joined #openstack-infra08:59
hasharlifeless: that is partly why I haven't upgraded Zuul yet :-]09:00
*** Ryan_Lane has quit IRC09:01
*** davidhadas_ has quit IRC09:01
*** tkammer has quit IRC09:01
*** markmc has joined #openstack-infra09:01
*** DinaBelova has quit IRC09:08
openstackgerritlifeless proposed a change to openstack-infra/gerritlib: Allow ctrl-Cing gerritlib programs.  https://review.openstack.org/4933309:17
*** michchap_ has joined #openstack-infra09:26
*** tkammer has joined #openstack-infra09:27
*** tvb has joined #openstack-infra09:27
*** Ryan_Lane has joined #openstack-infra09:27
*** michchap has quit IRC09:29
*** dizquierdo has left #openstack-infra09:30
*** flaper87 is now known as flaper87|afk09:31
*** tvb has quit IRC09:31
*** Ryan_Lane has quit IRC09:32
*** AlexF has joined #openstack-infra09:32
ttxgrmbl, fail at top of queue after running the test for 2 hours09:33
ttxso it's not just flaky tests. That one took 2 hours to run (and fail): https://jenkins01.openstack.org/job/gate-tempest-devstack-vm-full/17906/09:36
ttxand then another hour to get the failure noticed and processed.09:37
ttx...and counting09:41
ttxJenkins is not properly returning a FAIL for that job, so the whole queue is stuck09:41
*** flaper87|afk is now known as flaper8709:42
ttxYep, all results on jenkins01 are stuck09:42
ttxWe lost it09:43
openstackgerritAntoine Musso proposed a change to openstack-infra/jenkins-job-builder: missing parameter error are no more meaningfull  https://review.openstack.org/4933609:44
ttx#status notice One of our Jenkins masters is failing to return results, so the gate is currently stuck.09:46
openstackstatusNOTICE: One of our Jenkins masters is failing to return results, so the gate is currently stuck.09:46
ttx#status notice Jenkins01 is not failing, it's just very slow at the moment... so the gate is not completely stuck.09:54
openstackstatusNOTICE: Jenkins01 is not failing, it's just very slow at the moment... so the gate is not completely stuck.09:54
ttxNote: the slowness apparently caused OK tests to fail: http://logs.openstack.org/29/49029/1/gate/gate-tempest-devstack-vm-full/7355145/console.html09:55
*** Ryan_Lane has joined #openstack-infra09:58
*** DinaBelova has joined #openstack-infra10:00
*** Ryan_Lane has quit IRC10:02
*** tkammer has quit IRC10:04
*** darraghb has joined #openstack-infra10:05
markmcttx, oh dear10:08
*** davidhadas_ has joined #openstack-infra10:09
ttxThere are test jobs like gate-cinder-docs that have been running for 1h22min, waiting for jenkins01 to notice they completed 1h20 min ago10:09
ttx(top-left on zuul status page)10:10
ttxerr.. actually that job is still running, which is even more puzzling.10:11
*** pblaho has quit IRC10:11
ttxLots of proxy errors to jenkins01 too10:12
*** davidhadas has quit IRC10:12
*** pblaho has joined #openstack-infra10:12
*** matsuhashi has quit IRC10:16
*** tkammer has joined #openstack-infra10:17
*** SergeyLukjanov has quit IRC10:27
*** schwicht has joined #openstack-infra10:27
*** tvb has joined #openstack-infra10:28
*** dizquierdo has joined #openstack-infra10:28
*** Ryan_Lane has joined #openstack-infra10:28
*** AlexF has quit IRC10:30
*** dims has joined #openstack-infra10:30
*** pblaho has quit IRC10:30
*** tvb has quit IRC10:32
*** Ryan_Lane has quit IRC10:32
*** SergeyLukjanov has joined #openstack-infra10:33
*** pcm_ has joined #openstack-infra10:33
*** pcm_ has quit IRC10:34
*** pcm_ has joined #openstack-infra10:34
*** yaguang has joined #openstack-infra10:35
*** dizquierdo has quit IRC10:36
*** AlexF has joined #openstack-infra10:38
*** sdake_ has quit IRC10:40
*** sdake has quit IRC10:41
*** sdake_ has joined #openstack-infra10:41
*** sdake_ has joined #openstack-infra10:41
*** sdake has joined #openstack-infra10:41
ttxand/or scp being dead slow10:44
*** jeblair has quit IRC10:47
*** jeblair has joined #openstack-infra10:48
*** AlexF has quit IRC10:50
*** weshay has joined #openstack-infra10:53
*** AlexF has joined #openstack-infra10:55
*** Ryan_Lane has joined #openstack-infra10:58
*** fbo is now known as fbo_away10:59
*** dizquierdo has joined #openstack-infra11:00
*** johnthetubaguy1 is now known as johnthetubaguy11:01
openstackgerritEndre Karlson proposed a change to openstack-infra/config: Add RTFD hook to trigger auto-build of docs  https://review.openstack.org/4843711:01
*** alexpilotti has joined #openstack-infra11:02
*** Ryan_Lane has quit IRC11:03
*** sdake_ has quit IRC11:03
*** sdake_ has joined #openstack-infra11:03
*** sdake_ has joined #openstack-infra11:03
*** markmc has quit IRC11:04
openstackgerritEndre Karlson proposed a change to openstack-infra/config: Add RTFD hook to trigger auto-build of docs  https://review.openstack.org/4843711:05
*** darraghb has left #openstack-infra11:06
*** SergeyLukjanov has quit IRC11:08
*** dims has quit IRC11:09
*** SergeyLukjanov has joined #openstack-infra11:10
*** yaguang has quit IRC11:12
*** adalbas has joined #openstack-infra11:15
*** nosnos has quit IRC11:16
ttxyes, bad karma confirmed for jenkins01: http://cacti.openstack.org/cacti/graph_view.php?action=tree&tree_id=1&leaf_id=4011:16
ttxstarting around 1900 UTC yesterday11:17
jd__damn :(11:19
*** julim has joined #openstack-infra11:19
*** afazekas_no_irq has quit IRC11:20
*** luhrs1 is now known as che-arne11:21
sdaguehmmmm11:23
*** tvb has joined #openstack-infra11:23
sdaguealso this is an interesting thing - https://review.openstack.org/#/c/49006/11:23
sdaguethe change was ninja merged by mordred11:23
sdaguebut garyk ran reverify11:23
sdagueso now it's an unpassable change in the gate11:23
sdagueand because it's already merged, my normal ninja "new review" trick can't snipe it out11:24
*** CaptTofu has quit IRC11:25
*** CaptTofu has joined #openstack-infra11:26
*** Ryan_Lane has joined #openstack-infra11:29
*** Ryan_Lane has quit IRC11:33
ttxsdague: I think jenkins01 weird condition is the top fire atm11:35
ttxsdague: creates lots of slowness, superfluous fails and edge behavior11:35
*** dims has joined #openstack-infra11:36
ttxsdague: looks like we could use an alarm on that. It's been burning CPU for the last 16 hours now11:37
ttxgoing slightly better now as jenkins02 is handling most of the load, but some jobs are still getting on jenkins01 tarpit11:39
*** dims has quit IRC11:41
*** dims has joined #openstack-infra11:42
*** sandywalsh has quit IRC11:49
*** flaper87 is now known as flaper87|afk11:53
*** CaptTofu has quit IRC11:54
*** CaptTofu has joined #openstack-infra11:54
*** Ryan_Lane has joined #openstack-infra11:59
*** derekh has quit IRC11:59
*** fbo_away is now known as fbo12:01
*** Ryan_Lane has quit IRC12:03
ttxjenkins01 seemes to have recovered a bit12:05
ttxhttp://cacti.openstack.org/cacti/graph.php?action=view&local_graph_id=816&rra_id=all12:06
ttxmust be smelling that its masters are about to wake up12:07
ttxwe should be hiring a servant that doesn't get drunk all night while not being supervised12:08
*** AlexF has quit IRC12:08
AJaegerttx, just hire somebody in a different timezone ;)12:08
*** dprince has joined #openstack-infra12:09
*** DinaBelova has quit IRC12:09
jd__good point :)12:10
*** hashar has quit IRC12:12
*** mriedem has joined #openstack-infra12:12
*** OlivierSanchez has quit IRC12:16
*** flaper87|afk is now known as flaper8712:17
mriedemmordred: thanks a bunch! https://review.openstack.org/#/c/49006/12:18
*** dcramer_ has quit IRC12:21
*** hashar has joined #openstack-infra12:26
*** pblaho has joined #openstack-infra12:27
*** hashar_ has joined #openstack-infra12:29
*** pblaho has quit IRC12:29
*** thomasm has joined #openstack-infra12:30
*** Ryan_Lane has joined #openstack-infra12:30
*** pblaho has joined #openstack-infra12:30
*** osanchez has joined #openstack-infra12:31
*** hashar has quit IRC12:32
*** hashar_ is now known as hashar12:32
*** Ryan_Lane has quit IRC12:34
*** pblaho has quit IRC12:34
*** matty_dubs|gone is now known as matty_dubs12:35
*** markmcclain1 has joined #openstack-infra12:38
*** markmcclain has quit IRC12:38
*** dkranz has joined #openstack-infra12:42
openstackgerritDavid Caro proposed a change to openstack-infra/jenkins-job-builder: Added globbed parameters to the job specification  https://review.openstack.org/4868812:48
*** acabrera has joined #openstack-infra12:53
*** thomasbiege1 has joined #openstack-infra12:56
*** thomasbiege1 has quit IRC12:56
*** Ryan_Lane has joined #openstack-infra13:00
*** Ryan_Lane has quit IRC13:05
*** giulivo has joined #openstack-infra13:06
giulivoguys, I'm not really sure what is failing here?13:06
giulivohttps://review.openstack.org/#/c/47508/13:06
*** jcoufal_ is now known as jcoufal13:08
giulivoI don't think it is the tests but to make sure we address it properly with the recheck13:08
giulivojog0,13:09
ttxgiulivo: could be an artifact of jenkins01 being CPU-bound a few hours ago13:09
ttxgiulivo: I would "recheck no bug"13:09
ttx(in that precise case)13:09
giulivothanks ttx :)13:09
giulivojog0, I noticed people started to thank you for the hints about the bug they're hitting in the reviews :)13:10
*** markmcclain1 has quit IRC13:11
*** pblaho has joined #openstack-infra13:11
*** markmcclain has joined #openstack-infra13:12
*** markmcclain has left #openstack-infra13:12
*** markmcclain has joined #openstack-infra13:12
*** mrmartin has quit IRC13:13
sdaguefungi up yet?13:22
ttxfungi: up?13:22
*** changbl has quit IRC13:22
ttxsdague: wow.13:22
sdagueheh13:22
ttxsdague: WE SHARE THE SAME BRAIN13:22
ttxAt 3:22 pm we both got some synapse deciding fungi is usually up at that time.13:23
*** weshay has quit IRC13:24
sdague:)13:25
*** acabrera has quit IRC13:25
*** ryanpetrello has joined #openstack-infra13:25
*** dkehn_ is now known as dkehn13:26
*** cppcabrera has joined #openstack-infra13:27
mriedemi'm trying to figure out why this failed, but it's not obvious to me: http://logs.openstack.org/72/48772/1/check/check-tempest-devstack-vm-neutron/db02571/console.html13:27
mriedemdoes leaked process file descriptors in jenkins cause jobs to fail?13:27
*** michchap_ has quit IRC13:27
jd__mriedem: no13:28
*** alcabrera has joined #openstack-infra13:28
mriedemjd__: i didn't think so, otherwise nothing glaring as far as errors or failures13:28
*** bnemec has joined #openstack-infra13:28
jd__mriedem: is this really supposed to the end of the script at this stage?13:29
jd__I wonder if it didn't exited early13:30
*** thomasbiege1 has joined #openstack-infra13:30
mriedemjd__: that's what it looks like, anteaya was asking about the same last night in another patch13:30
mriedemjd__: this one: https://review.openstack.org/#/c/48971/13:31
*** Ryan_Lane has joined #openstack-infra13:31
mriedemit was just reverified and passed13:31
jd__it's possible, there has been issues with jenkins01 recently that could cause such failure IMHO13:31
jd__you should try a recheck13:31
mriedemok, in that other patch it was a different devstack job (grenade vs neutron) but the same kind of mystery, no errors, just exited early13:32
*** michchap has joined #openstack-infra13:34
*** yassine has quit IRC13:34
*** yassine has joined #openstack-infra13:35
mriedemhmm, thinking it's more than a fluke, failed here too: https://review.openstack.org/#/c/48300/13:35
mriedemboth are stable/grizzly patches, using quantumclient13:35
*** Ryan_Lane has quit IRC13:35
*** rfolco has joined #openstack-infra13:35
mriedemi don't even see logs for nova compute13:36
openstackgerritFelipe Reyes proposed a change to openstack-infra/jenkins-job-builder: Moved get_scenarios() function to a base module to make it easier to reuse  https://review.openstack.org/4926713:37
ttxsdague: in the gate queue, there is a failed change not triggering a reset: 49006,113:39
ttxnote sure how to read that13:40
ttxnot*13:40
*** DinaBelova has joined #openstack-infra13:41
mriedemhttps://bugs.launchpad.net/openstack-ci/+bug/123418113:45
uvirtbotLaunchpad bug 1234181 in openstack-ci "stable/grizzly patches are failing jenkins in check-tempest-devstack-vm-neutron" [Undecided,New]13:45
fungifinally caught up on scrollback. geez you people like to talk13:47
fungiand i find all the problems right at the end of the scrollback. lovely13:47
fungii need to find a way to read this stuff in reverse from now on13:48
* fungi checks to see if things are still broken13:49
fungijenkins01 seems to be running jobs okay at the moment. jenkins02 is being very slow to give me a webui13:50
fungifinally showed up (could have been at my end), and is also running jobs13:51
fungiat first glance everything seems to be okay--what are the currently observed issues, if any? or did they all self-correct?13:52
ttxfungi: you should read scrollback in reverse order13:52
fungittx: it feels like reading top-posted e-mail to me... breaks my brain somehow13:52
ttxfungi: so yeah, jenkins01 was REALLY SLOW and broke a number of things (extra failed tests, extra slowness)13:53
ttxSCP still seems a bit slow but no longer as much blocking13:53
* fungi digs through the cacti graphs on the jenkinses13:53
ttxfungi: still wondering about 49006,1 status at the gate13:54
ttxit's failed but not triggering any reset13:54
ttxbeen like this all day13:54
ttxfungi cacti graph at http://cacti.openstack.org/cacti/graph.php?action=view&local_graph_id=816&rra_id=all13:54
ttxfungi: was actually in a bad condition yesterday in your afternoon13:55
*** dcramer_ has joined #openstack-infra13:55
ttxsounds like we could use an alert on that one13:55
*** dkliban has joined #openstack-infra13:55
fungittx: note that mordred force-merged 49006 since it was for quantumclient (not neutronclient) and is as such untestable by the neutronclient jobs13:55
fungizuul just doesn't realize it's merged, and chucked it into the queue anyway13:55
ttxfungi: hmm, ok, will it just break stuff when at top of queue ?13:56
fungittx: that's a very good question, and one that i suspect i will have to wait to discover the answer for13:56
fungisince we do that so infrequently13:57
*** weshay has joined #openstack-infra13:58
*** thedodd has joined #openstack-infra13:59
*** DennyZhang has joined #openstack-infra13:59
fungii definitely find the cpu/load history on jenkins01 disturbing, and the memory utilization on jenkins0213:59
*** yassine has quit IRC14:00
ttxfungi: I can tell you it was pretty bad. Proxy errors trying to access it, then a docs job would take 2 hours to run just due to scp14:00
ttx(or appear so)14:00
*** yassine has joined #openstack-infra14:01
fungiyeah, clearly something was eating as much cpu (user) as it could get14:01
*** Ryan_Lane has joined #openstack-infra14:01
fungiwell, spiked to 80% user, looks like, and then plateaued there for about 14 hours14:02
*** thomasbiege1 has quit IRC14:03
fungiwhatever it was also seemed to free up a good 5gib ram when it cleared up too14:04
fungimaybe the start and/or end times of the plateau will loosely correlate to cron jobs on the machine14:05
*** thomasbiege1 has joined #openstack-infra14:05
*** Ryan_Lane has quit IRC14:06
*** AlexF has joined #openstack-infra14:07
ttxfungi: looks like classic jvm garbage collection metafail to me, but then I hate java14:08
*** thomasbiege1 has quit IRC14:09
fungiyeah, nothing suspiciously timed in various crontabs. i wonder if it has a monthly jvm cleanup task set to fire around 2100utc on the first of the month or something14:11
funginothing in the process list which seems to have restarted around the end of the plateau either14:12
*** AJaeger has quit IRC14:12
fungii'm starting to try and read documentation about profiling and tuning jvm garbage collection, and my brain is already turning to mush. this might require more coffee14:15
*** pblaho has quit IRC14:15
mriedemttx: fungi: just getting caught up too, https://review.openstack.org/#/c/49006/ was merged to unblock stable/grizzly patches14:21
mriedembut now i'm seeing at least 4 stable/grizzly patches all failing mysteriously in check-tempest-devstack-vm-neutron14:22
mriedemwhen setting up quantum14:22
jeblairttx: i've confirmed that the status page bug you observed is due to periodic jobs and is fixed by this patch from Alex_Gaynor: https://review.openstack.org/4931214:22
*** tvb has quit IRC14:22
*** markmc has joined #openstack-infra14:24
jeblairfungi: we lacked a number of graphs for jenkins0[12], including network; i just added them.14:24
ttxjeblair: cool.14:24
fungijeblair: yeah, i was wondering about that--what was the fix for the missing graphs? were they not being polled or just not in the config to be displayed?14:25
jeblairfungi: not being polled; it looks like when they were added, the snmp query to get interfaces/disks didn't run; so i manually triggered the queries and then ran the script again to create the graphs14:26
*** pblaho has joined #openstack-infra14:29
openstackgerritMatthew Treinish proposed a change to openstack-infra/elastic-recheck: Make test_required_files.py real unit tests  https://review.openstack.org/4926514:30
fungigot iy14:30
fungiit14:30
*** Ryan_Lane has joined #openstack-infra14:32
*** Guest49600 is now known as oubiwann_14:32
*** resker has quit IRC14:32
*** pblaho has quit IRC14:33
*** ruhe has joined #openstack-infra14:33
*** ryanpetrello has quit IRC14:35
*** ryanpetrello has joined #openstack-infra14:35
ttxjeblair: for some reason just before you get up, everything starts to behave properly again14:36
*** ryanpetrello has quit IRC14:36
*** Ryan_Lane has quit IRC14:36
fungii believe it senses his impending presence14:36
*** ryanpetrello has joined #openstack-infra14:37
*** rcleere has joined #openstack-infra14:37
ttxfungi: I suspect Jenkins is drinking all night long to forget about the daylight abuse14:38
*** dkliban has quit IRC14:38
fungicertainly a plausible explanation14:39
openstackgerritA change was merged to openstack-infra/elastic-recheck: Make test_queries save stdout in tox testr  https://review.openstack.org/4908014:40
openstackgerritA change was merged to openstack-infra/elastic-recheck: Add mox fixture to base TestCase  https://review.openstack.org/4926414:40
openstackgerritA change was merged to openstack-infra/elastic-recheck: Cleanup README  https://review.openstack.org/4928114:40
openstackgerritA change was merged to openstack-infra/elastic-recheck: Add query for bug 1232748  https://review.openstack.org/4928014:40
uvirtbotLaunchpad bug 1232748 in nova "[Postgres] DB error: (OperationalError) could not translate host name "localhost" to address" [Undecided,Confirmed] https://launchpad.net/bugs/123274814:40
*** ryanpetrello has quit IRC14:40
*** DinaBelova has quit IRC14:41
*** ryanpetrello has joined #openstack-infra14:41
*** anteaya has joined #openstack-infra14:42
*** dkehn_ has joined #openstack-infra14:42
openstackgerritFelipe Reyes proposed a change to openstack-infra/jenkins-job-builder: Added support for Git shallow clone parameter  https://review.openstack.org/4926814:43
*** changbl has joined #openstack-infra14:44
*** dkehn has quit IRC14:45
*** pblaho has joined #openstack-infra14:48
*** changbl has quit IRC14:48
openstackgerritFelipe Reyes proposed a change to openstack-infra/jenkins-job-builder: Added support for Git shallow clone parameter  https://review.openstack.org/4926814:49
*** rnirmal has joined #openstack-infra14:50
*** datsun180b has joined #openstack-infra14:51
*** AlexF has quit IRC14:51
*** odyssey4me has quit IRC14:56
*** thomasbiege1 has joined #openstack-infra14:56
*** tkammer has quit IRC15:00
*** changbl has joined #openstack-infra15:00
*** Cyril2 has quit IRC15:01
*** changbl has quit IRC15:01
*** dkliban has joined #openstack-infra15:02
*** Ryan_Lane has joined #openstack-infra15:02
*** Ryan_Lane has quit IRC15:06
*** adalbas has quit IRC15:07
*** changbl has joined #openstack-infra15:08
*** Ryan_Lane has joined #openstack-infra15:09
openstackgerritYuriy Taraday proposed a change to openstack-infra/git-review: Remove shebang line from git_review/cmd.py  https://review.openstack.org/4938315:09
*** thomasbiege1 has quit IRC15:11
*** mrodden has quit IRC15:15
*** thomasbiege1 has joined #openstack-infra15:17
openstackgerritYuriy Taraday proposed a change to openstack-dev/hacking: Allow APL text to start on the first text column  https://review.openstack.org/4938715:23
*** esker has joined #openstack-infra15:23
*** tvb has joined #openstack-infra15:23
*** tvb has quit IRC15:23
*** tvb has joined #openstack-infra15:23
*** oubiwann_ is now known as oubiwann15:26
*** tvb has quit IRC15:30
*** Ryan_Lane has quit IRC15:31
*** jomara has joined #openstack-infra15:32
*** flaper87 is now known as flaper87|afk15:33
*** ruhe has quit IRC15:33
sdaguefungi: you up for a ninja merge?15:34
*** ruhe has joined #openstack-infra15:34
fungirebooting precise29 as it was also unreachable and exhibiting cpu lockups in the console15:35
fungisdague: sure15:35
sdaguedims: what's the review again?15:35
fungisdague: what are the circumstances for this one?15:35
sdagueneutron has too many race issues to land a python-neutronclient fix for one of them15:35
dimshttps://review.openstack.org/#/c/49210/15:35
sdaguebecause it can't pass tempest 4 times15:36
* fungi is looking it over15:36
openstackgerritMatthew Treinish proposed a change to openstack-infra/elastic-recheck: Update elasticRecheck imports in bot.py  https://review.openstack.org/4887815:37
dimsfungi, httplib2 has a bug running under eventlet which causes bad fd(s) as rpc and httplib2 stomp all over each other. simple fix was to cleanup the connection pool in httplib215:38
fungiso it looks like it's only failing the neutron/neutronclient-only jobs. if it does somehow contribute to making those even less stable, it's not likely to destabilize gating for any other projects at least15:38
ttxjeblair: weird optimization idea: when change 2 at the gate is all successful (while change 1 is still running tests), would it make sense to merge both immediately ?15:38
ttx(see current state at the zuul status)15:39
ttx(the trick being, they sometimes don't run the same tests)15:39
jeblairttx: change 1 could still fail an unrelated test, like a unit test15:39
jeblairttx: exactly :/15:39
dimsfungi, right.15:40
ttxjeblair: yes, so it would have to be extra smart. Like only assume test success on tests it actually ran15:40
ttxyay, looks like we can have glance rc1 now15:40
*** pblaho has quit IRC15:40
ttxjeblair: just thinking out loud, igore me15:40
fungidims: can we get at least one neutron core to +2 that first?15:41
*** flaper87|afk is now known as flaper8715:41
fungii would feel a little less hesitant to force a patch into their project if they were complicit15:41
openstackgerritA change was merged to openstack/requirements: Raise Babel requirements to >= 1.3  https://review.openstack.org/4873915:41
*** dizquierdo has quit IRC15:41
*** dizquierdo has joined #openstack-infra15:42
mriedemmordred: does your fancy bot cascade this out now? ^15:42
jeblairmriedem: yeah, it should be automatic (it's a jenkins job, it'll run in a few mins)15:42
dimsfungi, there was +2 from Salvatore Orlando yday15:43
fungimriedem: it should open or update the requirements patches for all projects subscribed to them in a bit15:43
anteayattx yay15:43
fungidims: ahh, i see you only updated the commit message since then, which is what cleared his +2. good enough for me15:43
dimsthanks fungi15:44
jog0jeblair fungi: can someone restart elastic-recheck15:45
jog0it got borked due to some issues on ElasticSearch being slow and the fix patches aren't in yet15:45
openstackgerritMatthew Treinish proposed a change to openstack-infra/elastic-recheck: Reorganize tests into unit and functional tests  https://review.openstack.org/4939115:46
jeblairjog0: will do15:46
jog0jeblair: thanks15:47
jeblairjog0: done15:47
openstackgerritAnita Kuno proposed a change to openstack-infra/puppet-dashboard: Changing the module to use the sodabrew fork of puppet-dashboard  https://review.openstack.org/4939215:47
*** AlexF has joined #openstack-infra15:48
fungidims: sdague: merged15:48
*** oubiwann has quit IRC15:49
dimsfungi, sdague thanks!15:49
*** jcoufal has quit IRC15:49
openstackgerritJoe Gordon proposed a change to openstack-infra/elastic-recheck: Stop using While True to wait for ElasticSearch  https://review.openstack.org/4876615:49
*** AlexF has quit IRC15:50
*** marun has quit IRC15:50
jog0jeblair: it was related to console.html file being cut off15:50
*** Ryan_Lane has joined #openstack-infra15:50
hashar!launchpad nodepool15:50
openstackhashar: Error: "launchpad" is not a valid command.15:50
hasharargh15:50
*** ruhe has quit IRC15:50
*** thingee_zzz is now known as thingee15:51
*** yassine has quit IRC15:52
*** boris-42 has quit IRC15:54
*** mrodden has joined #openstack-infra15:55
hasharjeblair: hey :]  I can't report bug against nodepool! https://launchpad.net/nodepool15:55
hashar[Report a bug] is gray15:55
jeblairhashar: that's because it's impossible for there to be bugs! :)15:56
jeblairhashar: i'll set that up real quick, sorry.15:56
*** prad has joined #openstack-infra15:56
hasharit is ok15:56
jeblairhashar: also, i know i still have a lot to do with nodepool -- like write docs.  :(15:56
hashari don't even have a gear backed Zuul ...15:56
*** thomasbiege1 has quit IRC15:57
hasharanyway I noticed nodepool/myjenkins.py  which had methods to play with Jenkins nodes.  Wanted to request that to be upstreamed.   Might do it myself though15:57
jeblairhashar: yeah, i've been meaning to do that too.  if you get to it first, that would be great.15:58
jeblairhashar: bug tracker configured15:59
hasharjeblair: it is all about finding out the original authors + reading the bzr man page15:59
*** Ryan_Lane has quit IRC15:59
hasharawesome15:59
fungijeblair: clarkb: per the thread on the infra list, the fuel devs are wanting to rename several of their projects. objections to adding them to the batch on saturday if they're ready in time?15:59
jeblairfungi: wfm, thx.16:00
fungii'll reply to the list with the details then16:00
fungialso, precise29 is back online in jenkins and successfully running jobs after the reboot16:01
*** gyee has joined #openstack-infra16:03
*** hogepodge has joined #openstack-infra16:03
*** hemnafk is now known as hemna_16:04
*** oubiwann has joined #openstack-infra16:06
*** portante is now known as portante|afk16:06
*** AlexF has joined #openstack-infra16:07
*** prad has left #openstack-infra16:09
anteayawelcome back precise2916:10
*** markmc has quit IRC16:10
*** hogepodge has quit IRC16:10
*** matty_dubs is now known as matty_dubs|lunch16:12
*** jpich has quit IRC16:12
*** changbl has quit IRC16:12
*** hogepodge has joined #openstack-infra16:12
*** AlexF has quit IRC16:13
hogepodgepabelanger: ping16:14
russellbdang this new zuul scheduler is the hotness16:14
jeblairoh thank goodness someone likes it16:15
russellbheh, haters gonna hate16:16
russellbi'm an infra fanboy i think16:16
notmynamerussellb: how can one not be?16:16
russellbexactly16:16
notmynamerussellb: and then you say, "I know, I'll work on a non-openstack project". now you have code and ci problems :-)16:17
russellbusually complete-lack-of-ci problems, heh16:17
notmynameya :-)16:18
fungian absolute vacuum of ci16:18
fungiechoing sounds16:18
pabelangerhogepodge, pong16:19
hogepodgeIs there much more I should do on https://review.openstack.org/#/c/49020/16:20
*** SergeyLukjanov has quit IRC16:20
openstackgerritJames E. Blair proposed a change to openstack-infra/elastic-recheck: Add hits_by_query method  https://review.openstack.org/4939816:20
openstackgerritJames E. Blair proposed a change to openstack-infra/elastic-recheck: add check_success tool  https://review.openstack.org/4924616:20
*** hashar has quit IRC16:20
pabelangerhogepodge, nope, I think we are good.16:22
pabelangerHopefully fungi or clarkb can approve16:22
hogepodgeprobably swamped with release 2 weeks away :-)16:22
jeblairhogepodge: we have 2 more weeks of this!  yikes.16:23
fungiyeah, it's been insane. somehow i feel like i've been completely busy and also accomplished absolutely nothing (or if i did, can't remember at all what it was)16:23
fungiseems somehow substantially busier than 6 (much less 12) months ago16:24
hogepodgeHK feels bigger16:24
jeblairpabelanger, hogepodge: left a comment; based on the commit msg, i think you want to remove the lines i pointed to, but i may be missing context.16:25
hogepodgejeblair: targets master by default? so not necessary?16:26
hogepodgethere could be a branch in the future to handle stable/(grizzly|havana)16:26
hogepodgeBut for now the module is simple enough to be universal.16:27
jeblairhogepodge: i would expect a future branch to be able to run those tests, so no qualification is necessary16:27
jeblairhogepodge: (what's currently there means "only run this job on master")16:28
jeblairhogepodge: (which is redundant as long as the only branch is master)16:28
hogepodgeI see. I'll pull it out. Thanks.16:28
clarkbfungi: iknowright!16:28
pabelangerdoh, nice catch16:28
clarkbI am going to be mostly afk today. driving back to seattle land. weather is much better should be fun :)16:29
fungiclarkb: good luck16:29
clarkbjog0 I plan on doing the exponential.backoff thing tonight16:29
jeblairclarkb: sounds nice!16:29
hogepodgejeblair pebelanger maybe I can catch up with you at the summit and learn a bit more about the inner workings of the infra project16:29
anteayaclarkb: I hope you have a safe drive16:30
jeblairhogepodge: i would love nothing more than to share that with you and help you assimilate^Mcontribute!16:30
clarkb++16:30
jeblairhogepodge: in case you haven't seen it, http://ci.openstack.org/ has some info.  much of it is correct.16:30
openstackgerritJoe Gordon proposed a change to openstack-infra/elastic-recheck: Stop using While True to wait for ElasticSearch  https://review.openstack.org/4876616:31
jog0clarkb: !16:31
hogepodgejeblair been swimming in that for a week or so now :-)16:31
jog0enjoy the drive16:31
*** dmakogon_ipod has joined #openstack-infra16:32
clarkbnotmyname: I think I am going to have to quote you on that :)16:32
hogepodgejeblair I learn better by doing, so at some point I want to dive into http://ci.openstack.org/running-your-own.html16:35
*** dmakogon_ has joined #openstack-infra16:36
*** dizquierdo has left #openstack-infra16:36
fungihogepodge: also we're starting to curate our conference presentations at http://docs.openstack.org/infra/publications/ if you like slides16:36
openstackgerritJoe Gordon proposed a change to openstack-infra/elastic-recheck: Stop using While True to wait for ElasticSearch  https://review.openstack.org/4876616:37
*** dmakogon_ipod has quit IRC16:38
*** marun has joined #openstack-infra16:39
*** emagana has joined #openstack-infra16:39
jog0jeblair: I see you toook over patch 4924616:40
jog0my same comments from before still apply to it16:40
jeblairjog0: not really (at least not yet); i just needed a function from it, so i pulled out the function and rebased16:41
jog0ohh16:41
*** tvb has joined #openstack-infra16:41
jeblairjog0: sorry, i should leave a comment about that, will do real quick16:41
*** dmakogon_ has quit IRC16:42
*** dmakogon_ipod has joined #openstack-infra16:42
jog0just did it16:42
jog0where is your dependant patch?16:42
jeblairjog0: i'm writing it16:42
jog0ahh what will it do?16:42
jeblairjog0: make graphs.  :)16:43
jog0jeblair: awesome sauce16:43
jeblairyeah, i think it's going to be really useful16:43
jog0so one thing that would be neat is to show of all failures what percentege we have classified16:44
jog0will the graphs be time series as well?16:44
*** julim has quit IRC16:44
jeblairjog0: yes, i'm making a page of 'bug hits over time' for each bug.16:44
*** adalbas has joined #openstack-infra16:44
jog0cool, kibnana has that if you know what to search16:45
*** Ryan_Lane has joined #openstack-infra16:45
jog0which is to say people don't know what to search16:45
jeblairjog0: but that's still one bug at a time, right?16:45
*** tvb has quit IRC16:46
jog0jeblair: yes, its not that useful but maybe a way to use that as part of your solution16:46
jog0jeblair: for example bug 123040716:48
uvirtbotLaunchpad bug 1230407 in neutron "VMs can't progress through state changes because Neutron is deadlocking on it's database queries, and thus leaving networks in inconsistent states" [Critical,Confirmed] https://launchpad.net/bugs/123040716:48
jog0http://logstash.openstack.org/#eyJzZWFyY2giOiIgICAgQG1lc3NhZ2U6XCJMb2NrIHdhaXQgdGltZW91dCBleGNlZWRlZDsgdHJ5IHJlc3RhcnRpbmcgdHJhbnNhY3Rpb25cIiAgICAgQU5EIEBmaWVsZHMuZmlsZW5hbWU6XCJsb2dzL3NjcmVlbi1xLXN2Yy50eHRcIiIsImZpZWxkcyI6W10sIm9mZnNldCI6MCwidGltZWZyYW1lIjoiNjA0ODAwIiwiZ3JhcGhtb2RlIjoiY291bnQiLCJ0aW1lIjp7InVzZXJfaW50ZXJ2YWwiOjB9LCJzdGFtcCI6MTM4MDczMjQ3ODYwN30=16:48
openstackgerritA change was merged to openstack-infra/elastic-recheck: Update elasticRecheck imports in bot.py  https://review.openstack.org/4887816:48
*** afazekas_no_irq has joined #openstack-infra16:48
*** SergeyLukjanov has joined #openstack-infra16:49
jog0or http://logstash.openstack.org/#eyJzZWFyY2giOiIgICAgQG1lc3NhZ2U6XCJ0ZW1wZXN0LnNjZW5hcmlvLnRlc3RfbmV0d29ya19iYXNpY19vcHMgQXNzZXJ0aW9uRXJyb3I6IFRpbWVkIG91dCB3YWl0aW5nIGZvclwiICAgICBBTkQgQGZpZWxkcy5maWxlbmFtZTpcImNvbnNvbGUuaHRtbFwiIiwiZmllbGRzIjpbXSwib2Zmc2V0IjowLCJ0aW1lZnJhbWUiOiI2MDQ4MDAiLCJncmFwaG1vZGUiOiJjb3VudCIsInRpbWUiOnsidXNlcl9pbnRlcnZhbCI6MH0sInN0YW1wIjoxMzgwNzMyNTUxNDcxfQ==16:49
jog0shows 1224001 is mostly fixed16:49
jeblairjog0: *nod*, i'm patterning it after that, but i don't think i can use it directly for a number of reasons.  that's definitely the info i want to see.16:49
jeblairi should have something shortly.16:50
jog0jeblair: cool, yeah this page will be really really useful16:50
*** julim has joined #openstack-infra16:52
*** sarob has joined #openstack-infra16:53
*** tstevenson has joined #openstack-infra16:55
*** emagana has quit IRC16:55
*** dcramer_ has quit IRC16:55
*** hogepodge has quit IRC16:55
*** ryanpetrello_ has joined #openstack-infra16:57
*** ryanpetrello has quit IRC16:57
*** ryanpetrello_ is now known as ryanpetrello16:57
*** julim has quit IRC16:58
*** matty_dubs|lunch is now known as matty_dubs16:59
*** fbo is now known as fbo_away17:01
*** hogepodge has joined #openstack-infra17:02
*** davidhadas_ has quit IRC17:04
*** tstevenson has quit IRC17:04
jeblairjog0: http://imgur.com/iFNa6nS17:05
jog0jeblair: nice17:06
jog0can you add the query in there too17:06
jog0and the bug title17:06
*** sarob has quit IRC17:07
jeblairjog0: i was planning on adding the query (and linking it to logstash); we'd have to go to lp to get the bug title, yeah?17:07
*** sarob has joined #openstack-infra17:07
jog0yeah17:08
jeblairjog0: let's leave that for the next round then.  :)17:08
jog0link to query would be cool17:08
jog0jeblair: works for me17:08
openstackgerritChris Hoge proposed a change to openstack-infra/config: Adding puppet-vswitch project to Stackforge  https://review.openstack.org/4902017:08
fungii wonder if that will be more useful as a percentage of failures found in overall jobs run by the same name (though that gets hairy since the same error could appear in more than one similar yet distinct job)17:09
openstackgerritMatthew Treinish proposed a change to openstack-infra/elastic-recheck: Reorganize tests into unit and functional tests  https://review.openstack.org/4939117:09
*** DinaBelova has joined #openstack-infra17:09
fungibecause as it stands, an occurrence tally will rise and fall with overall job volume17:09
*** harlowja has joined #openstack-infra17:10
jeblairfungi: i think you're correct17:10
fungior is that y axis already a proportion rather than a linear count?17:11
jeblairfungi: nope, it's a histogram17:11
fungihard to tell17:11
fungih17:11
fungik17:11
*** dkranz has quit IRC17:12
*** sarob has quit IRC17:12
*** jerryz has joined #openstack-infra17:20
jeblairjog0: know how to make one of those base64 encoded urls using python?17:21
openstackgerritA change was merged to openstack-infra/elastic-recheck: Add hits_by_query method  https://review.openstack.org/4939817:22
*** emagana has joined #openstack-infra17:22
mtreinishjeblair: something like:  base64.urlsafe_b64encode(s)17:24
mtreinishjust guessing17:24
jeblairmtreinish: yeah, though i think the input for logstash/kibana is slightly different than the elasticsearch query17:24
mtreinishjeblair: I could definitely see it being that way17:25
jog0jeblair: not sure17:26
jeblairit looks like a pretty straightforward translation, shouldn't be a big deal.17:26
jog0probablly can dig up the kibana source code17:27
*** dkranz has joined #openstack-infra17:29
*** tvb has joined #openstack-infra17:30
jeblairgot it17:31
*** emagana has quit IRC17:32
*** sarob has joined #openstack-infra17:32
*** DennyZhang has quit IRC17:34
*** dkehn_ is now known as dkehn17:35
matty_dubsmordred: jeblair: I appreciate the help yesterday with looking at AskBot. FWIW I'm bringing in jomara to lend some of his Puppet expertise to this endeavor, so he might have some questions as well.17:35
ryanpetrellopotentially dumb question - is there a reason you can't $ pip install horizon from PyPI?17:36
*** sarob has quit IRC17:36
*** fbo_away is now known as fbo17:37
*** Ajaeger has joined #openstack-infra17:37
jomaramordred: hi! when he says expertise, he doesnt mean that17:37
*** melwitt has joined #openstack-infra17:39
*** tzumainn has joined #openstack-infra17:40
anteayajomara: any level of puppet experience is welcome, however you want to characterize it17:41
*** davidhadas has joined #openstack-infra17:42
*** yolanda has quit IRC17:44
*** DennyZhang has joined #openstack-infra17:44
jomaraanteaya: i will characterise it as, "can use sledgehammer to tear down wall"17:45
fungiryanpetrello: good question actually... we've generally only released client packages onto pypi, as server packages are expected to get installed from distribution vendors instead17:46
*** tzumainn has left #openstack-infra17:46
ryanpetrelloyea, figured17:46
fungiryanpetrello: ...or potentially from source if you're doing continuous deployment17:46
ryanpetrellomy predicament is that I'm writing a Horizon extension for personal use, and need to set up tox to install *just* Horizon17:46
ryanpetrelloI suppose I could point at Github...17:46
fungiryanpetrello: we do currently publish release tarballs of the server projects on launchpad.net17:47
*** flaper87 is now known as flaper87|afk17:48
fungiryanpetrello: and if you need branch-tip tarballs those can be found on tarballs.openstack.org17:48
ryanpetrelloyep, that's probably the route I should go (the release tarball)17:48
fungipip will deal with those tarballs fine since they're already sdist'ed. if you're wanting to get source from github.com for some reason, DO NOT use their tarball feature17:49
pleia2jomara: I think that's how everyone uses puppet :)17:49
ryanpetrellofungi: yep, http://tarballs.openstack.org/horizon/horizon-master.tar.gz is probably what I want, then17:50
fungiryanpetrello: also, depending on your distro, you may simply be able to install their horizon release packages, but presumably either your distro is lacking there or you've already ruled that option out for other reasons17:50
ryanpetrellosure17:50
ryanpetrellothe trickiness is that I'm just installing into a virtualenv via tox17:50
ryanpetrelloso I can run tests against my code that uses horizon server17:50
ryanpetrellobut a direct tarball link should work17:50
fungiryanpetrello: yes, .*-master.tar.gz get rebuilt on every change merged to a project, so pretty close to cd17:50
ryanpetrellogreat - thanks!17:51
openstackgerritJoe Gordon proposed a change to openstack-infra/elastic-recheck: Cleanup tests  https://review.openstack.org/4876417:52
fungijomara: my sledge hammer is fairly worn at this point17:52
*** boris-42 has joined #openstack-infra17:52
anteayajomara: duly filed as such17:54
mriedemif i want to change something in the check-tempest-devstack-vm-neutron job, is that in devstack or devstack-gate?17:55
anteayaso it appears we could do with a fresh sledge, wonderful17:55
Ajaegerjeblair, fungi: any words of wisdom on this one https://review.openstack.org/#/c/48128/ ? I'm not sure which way to go...17:55
anteayaand a popular choice17:56
*** sgordon has joined #openstack-infra17:56
jeblairAjaeger: so i understand, why did you want 4 jobs?  and if you reduced to 2 jobs, what would they be?17:56
fungiAjaeger: yeah, so it seems like some of those tests being in separate jobs is simply so they'll get run in parallel to increase throughput/decrease overall run time17:57
Ajaegerjeblair: we have four different tasks we do: Syntax check, check for deletions and building of books - and a niceness check (whitespace)17:58
AjaegerWe can run them in parallel if we split them up.17:58
jeblairAjaeger: do they take long?17:58
*** changbl has joined #openstack-infra17:58
fungibut there are also potentially other ways to do that without having them all run on separate machines (for example if the scripts could be run in parallel on the same machine and then results collated)17:58
AjaegerOn the other hand, the build job is very long (2 mins for one book, 10 for all), the others are small (seconds)17:58
Ajaegerbuilding is done already in parallel on one machine17:59
*** julim has joined #openstack-infra18:00
AjaegerI can merge the three voting jobs into one, no problem at all. We (-docu) just considered it better to have one gate for each major test we do.18:00
*** flaper87|afk is now known as flaper8718:01
Ajaegerfungi, jeblair I don't think it makes much difference whether its 4 or 2 timewise.18:01
openstackgerritDavid Caro proposed a change to openstack-infra/reviewstats: Added some options and pep8 compliance  https://review.openstack.org/4940618:01
jeblairi'm okay with 4, especially if they represent distinct tests...18:03
Ajaegerjeblair: 4 tasks - but run from the same python script18:03
openstackgerritSergey Lukjanov proposed a change to openstack-infra/config: Fix fuel repos layout  https://review.openstack.org/4940718:04
openstackgerritDavid Caro proposed a change to openstack-infra/reviewstats: Added some options and pep8 compliance  https://review.openstack.org/4940618:05
*** sarob has joined #openstack-infra18:05
jeblairAjaeger: yeah, it sort of sounds like how we run several things (flake8, unit tests) from tox; those are separate jobs18:06
jeblairAjaeger: so that's why i think it's fine; but here's a test you can use to help decide --18:06
jeblairAjaeger: think about how the results will be reported to the user; whether or not it would help to have a separate line for that job in the report on gerrit18:07
*** sarob has quit IRC18:09
openstackgerritJames E. Blair proposed a change to openstack-infra/elastic-recheck: Add graph script  https://review.openstack.org/4941018:09
jeblairjog0: ^ part 1/218:10
openstackgerritJames E. Blair proposed a change to openstack-infra/elastic-recheck: Add graph script  https://review.openstack.org/4941018:10
fungiAjaeger: jeblair: yep, as long as there's a reason for job proliferation (particularly when it comes to improving usefulness of developer feedback), i'm good with it18:10
mtreinishjeblair: cool18:11
jeblairAjaeger: i think you should remove the failure and success messages for the niceness check -- they don't indicate anything out of the ordinary18:11
Ajaegerannegentle_, sgordon: What do you think? I'm not 100 % convinced that having them as four jobs will *always* help.18:11
mtreinishjeblair: I'm wondering how many of the hypothetical todos I put in the summit session description will be done by summit18:12
Ajaegerjeblair: ok, will do.18:12
jeblairmtreinish: ha, i haven't read it.  :)18:12
jeblairmtreinish: i just felt like i kept wanting to run queries to look at graphs and realized that wasn't the best use of everyone's time.  :)18:12
fungiAjaeger: i think as long as you feel like it will help sometimes, that's good enough. no need to burn more engineering time on an otherwise excellent review18:13
openstackgerritAndreas Jaeger proposed a change to openstack-infra/config: Create new gates for openstack-manuals  https://review.openstack.org/4812818:13
Ajaegerfungi: YEs, I think it will help sometimes.18:14
fungiwfm18:14
Ajaegerjeblair: I removed the the failure/success message and send a new version ^18:14
Ajaegerjeblair, fungi : While I have your attention: sgordon mentioned that you had some other comments on how to improve the docs jenkins jobs. Is there anything you like me to look at?18:15
*** osanchez has quit IRC18:16
*** pblaho has joined #openstack-infra18:16
fungiAjaeger: well, the biggest thing for me is that many of them build with maven even though they're not java programs... seems a little odd, but probably a pragmatic choice at some point in the past which has not been reevaluated lately?18:17
fungiAjaeger: also, the way a lot of the documentation shares common target locations makes those places more likely to accumulate cruft since there's no safe way to tell which files are safe to remove18:18
Ajaegerfungi: regarding maven: no clue about that one18:19
Ajaegerfungi: yes, the cruft is bad. We would need to run rsync --delete-after instead of uploading via ftp18:19
Ajaegerbut there's no publisher for it ;(18:19
jeblairAjaeger: can you propose a summit session to talk about infra needs for docs (if there isn't one already)?18:20
Ajaegerjeblair: I'm not coming to the summit ;(18:20
jeblairAjaeger: nooo!18:21
Ajaegerjeblair: next time ;)18:21
jeblairAjaeger: want to try to remote18:21
jeblairremote-participate, or is the timezone too bad for you?18:21
fungiAjaeger: well, yes there are technical challenges with cleaning up cruft now, but if the jobs aren't designed to provide any hope of making that possible then there won't be a lot of impetus to solve the technical issues (different publisher mechanisms, maybe different hosting systems)18:21
* Ajaeger needs some time to see whether anything can be done here...18:23
Ajaegerperhaps some kind of filelist to iterate over?18:23
Ajaegerjeblair: let me check what's possible remote, hadn't thought about it18:24
fungiAjaeger: refactoring the manuals into separate git repositories with another repository for the theming/template pieces might also be helpful. right now a typo in one manual means we rebuild and reupload 20 other manuals18:24
Ajaegerfungi: indeed18:24
fungiand extracting out templates/theme bits also makes them more generally reusable elsewhere18:25
Ajaegerthe books share common files (documentation files shared by guides) and some build tools.18:27
sgordonAjaeger, i think what fungi just described is what i was talking about when i said we build too much18:27
sgordonthe validate/build on the review only builds impacted stuff18:27
sgordonbut publishing builds it all18:27
Ajaegersgordon: agreed18:27
sgordonand yeah common content is an on-going issue18:28
Ajaegerfungi, jeblair thanks for the food for thought, will see what I can do...18:28
sgordoncould be handled via a git submodule though18:28
fungisgordon: no need for submodules there18:28
fungisgordon: the build job can easily grab code from more than one git repository18:29
Ajaegerfungi: still, it needs to be easy for us humans as well18:29
fungi(and also there's the issue that much of our ci tooling in its current form can't support submodules, nor do we want it to)18:29
sgordonyeah but most existing editors and validators wont18:29
sgordonso you screw the authors18:30
sgordonok so that probably explains how we ended up where we are with all the books in one repo alongside the common content folder ;)18:30
fungii'm sure my limited view comes from writing and editing content abstracted completely from presentation18:30
*** johnthetubaguy has quit IRC18:30
*** sdake has quit IRC18:30
*** sdake has joined #openstack-infra18:30
openstackgerritJames E. Blair proposed a change to openstack-infra/elastic-recheck: Add graph script  https://review.openstack.org/4941018:32
Ajaegerfungi: still it might help us look at it from a differentangle18:33
fungisgordon: another alternative to submodules which allows you to provide the common template/boilerplate/formatting pieces in-tree with the unique content is the approach which oslo-incubator uses18:34
jog0jeblair: I think elastic-recheck bot is hung again18:34
jog0its working locally on my box and found two more bugs that it didn't18:34
fungisgordon: basically keep a master copy of ommon files in a separate repository and have a script which copies that into the document to update it, then make sure reviewers insist that changes to the common parts go to the common repository18:35
*** dkranz has quit IRC18:36
Ajaegerfungi: is there a way to run only some gate jobs based on the output of some script?18:36
AjaegerWe have a script that figure out which books to build - but don't use it for the publishing jobs to limit which to run18:37
jog0sdague mtreinish: can I get some eyes on https://review.openstack.org/#/c/48766/18:37
sgordonfungi, the second half is actually the part i suspect we'll be able to handle sooner18:37
fungiAjaeger: there is a way to run specific jobs when changes touch files matching a list of patterns18:37
jog0it should make elastic-recheck more robust18:37
sgordoni have on my list to write a check that shared content is in the common directory (or more accurately that shared content does not exist outside of it)18:37
sgordonthe former is going to require more invasive work in clouddocs-maven which im not as familiar with18:38
fungiAjaeger: that way it might be possible to only build manual-x when only files in its directory are changed, but rebuild all manuals when files in the shared content directory change18:39
fungiAjaeger: but since i don't know a lot about how the source repository for the manuals is organized, i have no idea whether that will help18:39
Ajaegerfungi: let me investigate that one, there might be some lowhanging fruit here18:39
*** mrmartin has joined #openstack-infra18:40
Ajaegerfungi: do you have a pointer or example for this?18:41
jog0fungi: ^^^ elastic-recheck needs a light kick18:41
fungijog0: restarting it now18:41
*** portante|afk is now known as portante18:41
jog0fungi: thanks! not sure what happened it kept working fine for me18:41
fungiAjaeger: i'll get you a link momentarily18:41
fungijog0: restarted18:42
sdaguejog0: looking18:42
jog0sdague fungi thanks18:43
lifelessmordred: so I made git@github.com:rbtcollins/gerrit-hook.git18:43
sdaguejog0: sure, I think we need to fix it in a real way later, but I'm ok with the patch for now18:43
lifelessmordred: I want to have a release of it and put it in gerrit; is it ok to put it in openstack/ or should it go somewhere else?18:44
jog0sdague: part of a real fix is to move away from polling logstash18:44
jog0all together18:44
*** DinaBelova has quit IRC18:44
sdagueclarkb: so what are the odds we could add the cloud-region to the logstash queries?18:45
sdaguethat would be kind of interesting18:45
fungiAjaeger: the files job parameter is described in http://ci.openstack.org/zuul/zuul.html#jobs and you can see it in use at https://git.openstack.org/cgit/openstack-infra/config/tree/modules/openstack_project/files/zuul/layout.yaml#n22918:46
*** sarob has joined #openstack-infra18:46
Ajaegerfungi: thanks!18:46
Ajaegerwill bookmark and read later18:46
* Ajaeger needs to do some other stuff now...18:46
Ajaegerfungi: yeah, I see some low-hanging fruit!18:47
fungiAjaeger: pick it when it's ripe!18:47
openstackgerritA change was merged to openstack-infra/elastic-recheck: Stop using While True to wait for ElasticSearch  https://review.openstack.org/4876618:47
fungior maybe pick it early and let it ripen? i really don't know what idiom i was going for there18:47
Ajaegerfungi: will do either way ;) Thanks a lot!18:48
fungilifeless: we already have some of those in openstack-infra/jeepyb... could we replace them with your thing or break them out of jeepyb and add them into yours? if so, i'd say it fits well in the openstack-infra org18:50
*** jerryz has quit IRC18:51
*** dkranz has joined #openstack-infra18:51
jog0sdague: should be doable. ask clarkb when he gets back from driving18:53
openstackgerritJames E. Blair proposed a change to openstack-infra/elastic-recheck: Add graph script  https://review.openstack.org/4941018:54
sdagueman, sorting python dicts by value elements always gets me turned around18:54
lifelessfungi: I'm easy18:54
lifelessfungi: I just want a place to put it and do releases from18:54
lifelessfungi: if you say 'openstack-infra with openstack-infra-core as reviewers' thats fine, I'll submit the config patch for that18:55
*** dcramer_ has joined #openstack-infra18:56
*** vipul is now known as vipul-away18:57
fungilifeless: i think the deciding factor there is "if it's useful to the infra team or for users/devs interacting with openstack's developer community infrastructure"18:59
*** sarob has quit IRC19:00
fungiso probably, but i'd get a second core opinion just for confirmation19:00
*** sarob has joined #openstack-infra19:00
*** danger_fo_away is now known as danger_fo19:00
*** sarob has quit IRC19:01
openstackgerritJames E. Blair proposed a change to openstack-infra/config: Add elastic-recheck graphs to status  https://review.openstack.org/4941719:03
jeblairjog0: ^ change 2/219:03
jeblairjog0, sdague, mtreinish, clarkb: http://status.openstack.org/elastic-recheck/19:03
openstackgerritSean Dague proposed a change to openstack-infra/elastic-recheck: add check_success tool  https://review.openstack.org/4924619:04
hub_capwhat version of centos is py26 run on?19:05
jeblairhub_cap: CentOS release 6.4 (Final)19:06
jomarajeblair: hey, n ot sure if you saw matty dubs introduction, but he handed off the askbot thing to me and i need to get up to speed19:06
zulcoolio looks like we might be able to making python3 gate voting for python-novaclient soon19:06
hub_cap<3 jeblair19:06
jeblairjomara: how can i help?  (i'm about to grab lunch though, i may have to help you when i get back)19:07
*** DinaBelova has joined #openstack-infra19:08
*** changbl has quit IRC19:10
jomarajeblair: no problem, you can lunch on19:10
jomarajeblair: i have basically no details other than the goal i need to accomplish and a tiny bit of history, so i just know i need to perform the installation with puppet :)19:11
*** thingee is now known as thingee_zzz19:19
openstackgerritSean Dague proposed a change to openstack-infra/elastic-recheck: add check_success tool  https://review.openstack.org/4924619:22
jeblairlifeless: the name of your tool confused me...19:25
jeblairlifeless: gerrit has hooks, so i thought it was a tool to help you run things using a gerrit hook19:26
lifelessfungi: have you looked at the code? Were you similarly confused with jeblair ?19:26
jeblairlifeless: but instead, this watches the event stream and performs actions based on that19:26
fungilifeless: oh, i made assumptions based on the name/readme19:26
fungihadn't looked closely yet19:27
lifelessfungi: the readme is clear I think19:27
lifelessthe name I did not realise would be confusing19:27
fungii may have read into it what i expected, in that case19:27
jeblairlifeless: which reminds me that we have a tool that does something similar... have you considered using zuul + https://github.com/rcbau/turbo-hipster ?19:27
lifelessjeblair: yes, its more moving parts than I want to consider for now19:27
lifelessjeblair: since we /want/ serialisation19:27
fungilifeless: my confusion is that gerrit has hooks, more or less like git hooks, which are usually shell scripts it runs on specific gerrit events19:28
lifelessjeblair: a future iteration of the CD cloud may well move to zuul19:28
lifelessfungi: what context do they run in?19:28
fungilifeless: on the gerrit server19:28
jeblairlifeless: ok.  that's possible with the independent queue manager + single worker.  also, new queue managers are possible.19:28
lifelessok, so this is remote19:28
fungiyeah, we have gerrit hooks (it even explicitly calls them hooks) which we run on gerrit servers19:28
lifelessjeblair: yes, I'm aware it's possible, and am interested in that in future, but higher entry cost19:29
fungiso i read far too much into the name and let it cloud my interpretation of the description19:29
lifelessjeblair: which may just be my ignorance of the easiness of zuul19:29
lifelessso in some ways I'd like this to be on stackforge19:29
lifelessto indicate it's a thing but maybe it's not a Thing19:30
lifelessbut if it sstuck around I don't want you guys to pay the overhead of moving it in future19:30
fungiyes, stackforge seems more appropriate if its future is uncertain and we're not using it in the upstream infrastructure19:30
fungiand i agree this seems far more like something we'd use zuul for if we did need it in the upstream infrastructure19:31
*** sarob has joined #openstack-infra19:31
lifelessright, and I have zero interest in replacing zuul; this is totally about having a low cost quick-experiment facility19:32
lifelessjeblair: fungi: any suggestions for a better name?19:32
lifelessgerrit-stream-reactor?19:32
*** sarob has quit IRC19:32
fungidont-cross-the-streams19:32
*** sarob has joined #openstack-infra19:33
fungigerrit-eventerator19:33
* fungi sucks at names19:34
annegentle_gerrit-rivulet (new name for me, means a stream) :)19:35
* ttx closes shop for today19:36
jeblairlifeless: if it's a quick experiment with an uncertain future (actually, you seem to be saying it has a probable dead-end future), why put it in gerrit at all?19:36
jeblairttx: see http://status.openstack.org/elastic-recheck/ ?19:36
*** sarob has quit IRC19:37
*** julim has quit IRC19:37
lifelessjeblair: if it's not in gerrit, it's 10 times harder to collaborate on; if it doesn't have releases it's much harder for folk to reason about use of it19:37
lifelessjeblair: and we have a team of 20ish folk driving tripleo so those factors do matter to us19:38
*** sarob has joined #openstack-infra19:39
sdaguejeblair: so you might want to liberate some stuff from the check_success tool19:39
sdaguethrew some comments in your review19:39
jeblairsdague: well, i clearly liberated most of the structure from check_success.  thanks.  :)  want a co-authored-by line?19:40
sdagueheh19:40
sdaguewell I just did a refactor, and found some other issues :)19:40
sdagueyou might want to reliberate19:40
sdaguelike my set counting was error prone19:40
jeblairsdague: oh, ok, the uuid is for de-duplication then?19:41
sdagueyep19:41
jeblairsdague, jog0: oh, i thought you wanted entry points rather than standalone tools...19:41
sdaguejeblair: I think there is debate, I thought we were moving to a tools model19:42
jeblairsdague: i like entry points because, as you'll see in https://review.openstack.org/#/c/49417/ i can just install e-r, and then call the script from cron19:43
jeblair(without having to run it out of the repo)19:43
*** DinaBelova has quit IRC19:43
jeblairsdague: but i'll go with the flow if you guys don't want that.19:43
sdaguesure, would there be an issue to have the entry points over in tools?19:43
*** dkranz has quit IRC19:44
sdaguewell things are moving fast, so lets say there is creative collision :)19:44
lifelesshow about19:44
lifelessgerrit-client-hooks ?19:44
lifelessfungi: jeblair: ^19:44
jeblairsdague: i don't think tools would be installed...19:44
jeblairsdague: (since i think setup.cfg just says to install elastic_recheck)19:45
sdagueright, it does now. Sorry I hadn't thought about the entrypoints bit yet19:45
fungilifeless: it's at least as good as echo `shuf /usr/share/dict/words|grep '^[a-z]*$'|head -n$(($RANDOM%3+1))`|sed 's/ /-/g'19:45
lifelessfungi: high praise!19:45
mordredfungi likes giving out the praise19:46
jeblairlifeless: that is less confusing (it would at least make me think twice; i couldn't say for sure that i'd get the idea right away)19:46
lifelessfungi: more importantly, does it reduce the confusion with gerrit serverside hooks ?19:46
lifelessok cool19:46
lifelessI will rename it and submit a config request for it for stackforge19:46
fungilifeless: probably, yes19:46
sdaguejeblair: let me fix my patch19:47
sdaguels19:48
jeblairsdague: the graph script is designed to keep track of all job results, so i think it's okay that success shows up (as long as it's getting categorized properly)19:49
jeblairsdague: does that make sense, or am i missing something?19:49
sdaguesuccess showing up I think is fine19:49
sdaguebut I thought there was still a hit count issue, as you are just using results19:49
openstackgerritSean Dague proposed a change to openstack-infra/elastic-recheck: add check_success tool  https://review.openstack.org/4924619:50
jeblairsdague: yeah, i'll fix that19:50
sdagueok, so that does it with entry points, but also lives outside the elastic_recheck dir19:50
jeblairsdague: oh, that works?19:51
sdaguejeblair: yep, did for me19:51
jeblairneato19:51
jeblairdidn't expect that19:51
annegentle_jeblair: sdague: by the way, how do you do a co-authored line on a patch?19:53
jeblairannegentle_: Co-Authored-By: foo <foo@example.com>19:53
annegentle_jeblair: cool I had a need for that a bit ago19:54
mordredsdague: I don't think that's going to work as well as you think in a pure install19:54
mordredsdague: because tools.check_success is not going to get installed19:54
*** oubiwann has quit IRC19:55
sdaguemordred: ok19:55
jeblairsdague: elastic_recheck/cmd/check_success19:55
mordredsdague: so the entry point created for elastic-recheck-success will point to an unresolvable path. however, when you run develop and you're still in that dir, the entry point will find it19:55
mordredsorry19:55
*** vipul-away is now known as vipul19:55
jeblair(i think that's a convention i've seen elsewhere)19:55
sdaguemordred: ok, so if we move it to cmd I'm good with it19:55
annegentle_jeblair: does that "do" anything (adding co-authored-by)?19:55
sdagueI just want some separation, and I really do want to be able to run some of the stuff from in tree19:56
annegentle_jeblair: I'll add info to https://wiki.openstack.org/wiki/GitCommitMessages19:56
jeblairannegentle_: nope, but it's a convention, and we've been encouraging the stats tools to observe it19:56
jeblairannegentle_: i thought it was there!19:56
annegentle_jeblair: cool, just did a search in the wiki and that page and couldn't find it19:56
jeblairannegentle_: i agree it's not.  i'm confused.19:56
sdaguemordred: so where should I fully fill my head with entry point info, so I don't be silly in the future?19:57
mordredsdague: muh...19:58
mordredsdague: you really don't want to know more :)19:58
mordredsdague: actually, it's easy as implemented by pbr19:58
mordredsdague: what happens is that, for the thing listed as the target of the entrypoint19:58
sdagueok, I just do want to find a way to run from inside the tree easily as well19:58
*** adalbas has quit IRC19:58
mordreda very small script is generated that does "from %(target_lib)s import %(target_object)s ; def main(): return %(target_object)s"19:59
*** dcramer_ has quit IRC19:59
mordredsdague: so you could actually have a script in tools that just does that if the entry_point generation is causing you angst19:59
mordredthere is no other magic - it's just a way to avoid committing boilerplate code to tree at this point20:00
openstackgerritSean Dague proposed a change to openstack-infra/elastic-recheck: add check_success tool  https://review.openstack.org/4924620:00
sdaguemordred: ok, so better?20:01
mordredsdague: (if you look in /usr/local/bin, you should see the script that pbr generated - it should work unmodified if copied into tyour local dir)20:01
mordred>20:01
mordredI don't see this as about Lauren either. But, it is a real world example which I think is a useful benchmark. I worry we spend a huge amount of time anticipating all the bad things that *COULD* happen vs. what the reality is. I see a different analogy  here. It is not uncommon for a corporate BoD to have members of the company on the board. So, there are clearly examples of both options. It is up for20:01
mordredus to decide. I happen to think being more inclusive is better. So, I am in favor of allowing foundation employees on the board as suggested.20:01
*** julim has joined #openstack-infra20:01
mordredsdague: perfect20:01
mordredsdague: also, it looks like running "python elastic_recheck/cmd/check_success.py" should also work like a champ20:02
*** dmakogon_ipod has left #openstack-infra20:04
jomarajeblair: he mentioned something about there being an issue w/ setting up postgres20:04
jomarajeblair: is it just missing puppet classes?20:04
*** oubiwann has joined #openstack-infra20:05
jeblairjomara: for the full context, you might want to read http://eavesdrop.openstack.org/irclogs/%23openstack-infra/%23openstack-infra.2013-09-30.log starting at 2013-09-30T20:24:5820:05
*** tvb has quit IRC20:06
jomarajeblair: ahh, that will help, thanks20:06
*** dprince has quit IRC20:07
*** dkliban has quit IRC20:08
jog0jeblair: nice!20:08
openstackgerritAndreas Jaeger proposed a change to openstack-infra/config: Run some manual post jobs only if needed  https://review.openstack.org/4942920:08
Ajaegerfungi: here's a first ripe fruit20:09
jog0although the bar saying success of failure is a little long20:09
Ajaegersgordon: adding patterns for common + the directory should be easy once you cleaned up the references20:09
jog0jeblair: also can you make the bug numbers links to launchpad20:09
jeblairjog0: yeah, LOTS of things could be made better20:09
jog0jeblair: the links to launchpad is really simple20:10
jeblairjog0: there are links to launchad under the graph, next to the link to logstash20:10
jog0jeblair: just use 'https://bugs.launchpad.net/bugs/%s' % bug_number20:10
jog0ohh didn't see that nice20:10
jeblairi'd like to focus on structural issues and get this merged, then it'll be easier for others to hack on it20:11
jog0jeblair: ++20:11
fungijog0: you can even omit the bugs subdomain and launchpad.net's director will figure it out from the url pattern20:11
jog0fungi: hehe nice20:11
jog0bug 1224001 is impressive20:11
uvirtbotLaunchpad bug 1224001 in neutron "test_network_basic_ops fails waiting for network to become available" [Critical,In progress] https://launchpad.net/bugs/122400120:11
jog0in your graph jeblair20:11
jog0looks like it dropped off recently20:11
openstackgerritJames E. Blair proposed a change to openstack-infra/elastic-recheck: Add graph script  https://review.openstack.org/4941020:12
*** dcramer_ has joined #openstack-infra20:13
jeblairsdague: i did uuid de-duplication a little differently than you ^20:13
sdaguejeblair: ok, sure20:13
sdaguehonestly, next up is to write a class to handle the results in some more sane way. 5 levels of dictionaries, with leaves than might be strings or might be arrays is pretty ugly to carry around in the code20:14
jomarajeblair: i feel like this conversation is happening around me! ive given each of you voices20:14
*** julim has quit IRC20:15
mtreinishjog0: that drop off is because it only affects neutron now. sdague turned off tenant isolation off again for the neutron runs on everyone else20:15
jeblairjomara: we all have voices too!20:15
sdaguemtreinish is correct20:15
jog0mtreinish: oh right, its still neat to see though20:15
sdaguejeblair: I'd still vote sorting the graphs by failures20:16
jeblairsdague: yes, that (and many other things) would be nice, but i'd like to get this much merged so other people can start hacking/using it.20:17
jog0jeblair: ++20:18
sdaguejeblair: sure, I'm game20:19
jomarajeblair: it looks like the first thing proposed was , "run on openshift until we have time to setup a proper puppet install w/ postgres on openstack-infra"20:19
jomarathen migrate the db and put in a redirect or something20:20
jomarai think matty tried to mess wtih getting it to work w/ mysql but didnt have much luck, its probably not worth haxing up askbott20:21
mriedemdtroyer: sdague: fyi https://review.openstack.org/#/c/49409/20:23
mriedemthat's an attempt at fixing one of the common recheck bugs20:23
openstackgerritA change was merged to openstack-infra/elastic-recheck: add check_success tool  https://review.openstack.org/4924620:24
sdaguemriedem: I'm +2, if I don't hear from dtroyer before I leave in 30 mins, I'll +A it20:24
mriedemsdague: thanks20:24
sdaguethen it should get in overnight20:25
*** Ajaeger has quit IRC20:33
*** DennyZhang has quit IRC20:34
jog0jeblair sdague: about to merge https://review.openstack.org/#/c/49410/20:34
jog0jeblair: want to make sure you think its ready first20:35
jog0fungi: somethign is funny with the elastic-recheck bot running on infra20:36
jog0jeblair: ^20:36
jog0can we check the logs to see whats going on20:36
fungijog0: sure20:39
jog0fungi: a local copy just detected https://review.openstack.org/#/c/46979/220:39
jog0Found bug: https://bugs.launchpad.net/bugs/123274820:39
uvirtbotLaunchpad bug 1232748 in devstack "[Postgres] DB error: (OperationalError) could not translate host name "localhost" to address" [Undecided,In progress]20:39
openstackgerritlifeless proposed a change to openstack-infra/config: Add gerrit-client-hooks to our infrastructure.  https://review.openstack.org/4943220:40
openstackgerritlifeless proposed a change to openstack-infra/config: Better recommendations for adding sections.  https://review.openstack.org/4943320:40
openstackgerritlifeless proposed a change to openstack-infra/config: Better sample jobs.  https://review.openstack.org/4943420:40
*** dkranz has joined #openstack-infra20:41
lifelessIIRC we can do new things w/out downtime, right ?20:41
fungijog0: last entry in the debug log is 2013-10-02 20:40:55,033 DEBUG gerrit.GerritWatcher: Received data from Gerrit event stream20:41
fungilifeless: yes20:41
lifelessok, ^ plox :)20:41
lifelessbut I had to guess at the jobs list due to skew between practice and reality20:42
lifelesswhich I also captured as new docs.20:42
*** AlexF has joined #openstack-infra20:42
fungijog0: no mention of 46979 in the debug log20:42
jog0fungi: that may be misleading because gerritlib writes to logs and runs in the background20:42
jog0can you pastebin me the last 1000 lines or so?20:43
fungilast update to channel was at 19:4920:43
fungijog0: sure20:43
jog0fungi: TBH not sure what I am looking for either20:43
*** zul has quit IRC20:44
fungijog0: http://paste.openstack.org/show/4786720:44
fungijog0: oh, wait, you want debug. just a sec20:45
fungijog0: http://paste.openstack.org/show/4786820:45
*** SergeyLukjanov has quit IRC20:48
*** pblaho has quit IRC20:49
fungijog0: the git clone of elastic-recheck on logstash.o.o is also current with master20:49
*** sarob has quit IRC20:50
*** sarob has joined #openstack-infra20:51
jog0fungi: hmm no idea whats going on20:51
jog0fungi: can I get debug logs without gerritlib or irc?20:52
fungijog0: netstat on both ends seems to think the socket to gerrit is still open20:52
jog0last 1000 lines20:52
fungijog0: eh, probably not easily. there's way too much multiline stuff in that log20:52
jog0fungi: in that case last 10k lines in debug?20:52
fungicoming right up. or i can do a time range if you need20:53
jog0last two hours?20:53
*** alcabrera has quit IRC20:53
fungi10k lines is http://paste.openstack.org/show/47870 and goes back a little over 2h20:54
fungijog0: ^20:54
*** sarob has quit IRC20:55
openstackgerritJoe Gordon proposed a change to openstack-infra/elastic-recheck: Cleanup tests  https://review.openstack.org/4876420:57
fungijog0: filtered out all multiline continuations and gerritlib/irc client entries... http://paste.openstack.org/show/4787120:59
fungino idea if that view helps20:59
*** alexpilotti has quit IRC20:59
fungii suspect tweaks to the logging.conf might be in order to break the debug logs out in the future21:00
*** matty_dubs is now known as matty_dubs|gone21:00
*** datsun180b has quit IRC21:01
jog0fungi: going to add more debug logs21:01
*** freyes has quit IRC21:02
fungiokay21:02
*** alexpilotti has joined #openstack-infra21:07
openstackgerritJoe Gordon proposed a change to openstack-infra/elastic-recheck: Add more debug logs to classify  https://review.openstack.org/4944221:08
jog0fungi ^ that should help21:08
jog0sdague: can I get a quick review on ^21:08
openstackgerritDavid Caro proposed a change to openstack-infra/reviewstats: Added some options and pep8 compliance  https://review.openstack.org/4940621:10
dtroyersdague: +A in the localhost bit21:11
*** sarob has joined #openstack-infra21:12
*** changbl has joined #openstack-infra21:12
jog0fungi: if you can review my patch above we can rekick e-r after that lands21:13
fungiyou bet21:14
openstackgerritDavid Caro proposed a change to openstack-infra/reviewstats: Adapting for generic use  https://review.openstack.org/4944521:15
mtreinishjog0: are the logs from the bot posted somewhere?21:16
jog0mtreinish: fungi was doing it by hand21:17
jog0see scrollback21:17
*** prad has joined #openstack-infra21:17
*** jerryz has joined #openstack-infra21:19
*** dcramer_ has quit IRC21:19
*** sgordon has quit IRC21:21
zarofungi: would you be able to tell the purpose for gerrit::remotes module?21:22
*** thomasbiege1 has joined #openstack-infra21:24
fungizaro: we have some projects for which we are maintaining forks, so that updates branches containing the upstream versions21:25
*** thomasbiege1 has quit IRC21:26
openstackgerritA change was merged to openstack-infra/elastic-recheck: Add more debug logs to classify  https://review.openstack.org/4944221:26
zarofungi: thnx.21:26
*** mrmartin has quit IRC21:27
*** weshay has quit IRC21:28
fungijog0: restarted on top of that new commit now21:30
jog0fungi: thanks! hopefully that will give us a little more insight21:31
fungihere's hoping21:31
*** thomasm has quit IRC21:34
*** danger_fo is now known as danger_fo_away21:34
*** thingee_zzz is now known as thingee21:39
*** pcm_ has quit IRC21:45
openstackgerritlifeless proposed a change to openstack-infra/gerritlib: Allow ctrl-Cing gerritlib programs.  https://review.openstack.org/4933321:45
openstackgerritlifeless proposed a change to openstack-infra/gerritlib: Ignore editor files.  https://review.openstack.org/4932721:45
openstackgerritlifeless proposed a change to openstack-infra/gerritlib: Simplify GerritWatcher creation.  https://review.openstack.org/4932621:45
*** CaptTofu has quit IRC21:46
*** CaptTofu has joined #openstack-infra21:47
*** dkliban has joined #openstack-infra21:48
*** zul has joined #openstack-infra21:49
*** rfolco has quit IRC21:49
*** zul has quit IRC21:49
*** zul has joined #openstack-infra21:50
*** nati_ueno has joined #openstack-infra21:53
openstackgerritJames E. Blair proposed a change to openstack-infra/elastic-recheck: Add graph script  https://review.openstack.org/4941021:55
*** CaptTofu has quit IRC21:57
*** afazekas_no_irq has quit IRC21:57
*** CaptTofu has joined #openstack-infra21:58
openstackgerritA change was merged to openstack-infra/elastic-recheck: Add graph script  https://review.openstack.org/4941022:03
openstackgerritA change was merged to openstack-infra/config: Add elastic-recheck graphs to status  https://review.openstack.org/4941722:06
*** julim has joined #openstack-infra22:10
*** mriedem has quit IRC22:15
openstackgerritA change was merged to openstack-infra/config: Publish atom.xml for manuals  https://review.openstack.org/4789722:17
*** mrodden has quit IRC22:17
lifelessjeblair: would you like me to rebase my new project under the doc changes ?22:22
jeblairlifeless: that would be swell, thanks.22:22
openstackgerritKhai Do proposed a change to openstack-infra/config: decouple mysql setup from gerrit module  https://review.openstack.org/4944822:23
*** flaper87 is now known as flaper87|afk22:24
jeblairmordred, zaro: can the two of you get together and talk about https://review.openstack.org/#/c/48058/ and https://review.openstack.org/4532922:25
openstackgerritlifeless proposed a change to openstack-infra/config: Better sample jobs.  https://review.openstack.org/4943422:25
openstackgerritlifeless proposed a change to openstack-infra/config: Better recommendations for adding sections.  https://review.openstack.org/4943322:25
openstackgerritlifeless proposed a change to openstack-infra/config: Add gerrit-client-hooks to our infrastructure.  https://review.openstack.org/4943222:25
zaromordred: does 48058 mean that 45329 should be abandoned?22:28
zulwtf? why do we have to freaking update the requirements now i thought we were in a freeze?22:29
mordredzaro: nope. you keep yours. I was wrong about mine22:30
mordredzul: aroo?22:30
mordredzul: so - openstack/requirements is frozen - this is aligning the specific projects to that22:31
mordredzul: you should be happy about that, no?22:31
mordred(we did it for you)22:31
fungizul: which change? specifically?22:31
zulmordred/fungi: https://github.com/openstack/cinder/commit/56d3392e2dd66710a0876a8ba4d567e5d89086fb22:31
mordredzul: also, moving forward, they should drastically cease to diverge22:31
mordredzul: ah - that's a specific bug22:32
mordredzul: Bable 0.9.6 breaks with our current translations, 1.1 doesn't recognize zh_CN - we had to bump to 1.3 to get translation processing that actually worked with our repos22:32
zulmordred:  thats a really really big change for us at this close to release22:32
mordredzul: :(22:33
openstackgerritA change was merged to openstack-infra/publications: Update talk with new information and images.  https://review.openstack.org/4815122:33
mordredzul: lemme find the notes for you on what was broken22:33
dstufftis github showing notes a new thing? o.O22:33
mordredzul: https://review.openstack.org/#/c/48739/22:34
mordredzul: are the notes on it22:34
openstackgerritA change was merged to openstack-infra/publications: Add logstash slide.  https://review.openstack.org/4815222:34
*** prad has quit IRC22:35
mordredzul: since you're not installing via pip, you could probably ignore that, looking at it22:35
zulyeah22:35
zuljust get to carry another patch whee!!22:35
* zul goes get alcohol poising22:35
mordredzul: no, I mean, you should not need a patch22:36
mordredzul: you're doing SKIP_PIP_INSTALL in your rules file, yeah?22:36
openstackgerritA change was merged to openstack-infra/publications: Make zuul simulation reflect NNFI.  https://review.openstack.org/4815322:36
zulmordred:  SKIP_PIP_INSTALL?22:36
mordredzul: if you do that, then the contents of the requirements files are meaningless to you22:36
mordredzomg22:36
mordrednobody reads the docs I wrote for them22:36
mordredhttp://docs.openstack.org/developer/pbr/packagers.html22:37
mordredhttp://docs.openstack.org/developer/pbr/packagers.html#dependencies22:37
zulmordred:  sorry we dont ;)22:37
zulmordred:  i wasnt aware of them22:37
mordredzul: put SKIP_PIP_INSTALL in your debian/rules files22:37
*** oubiwann has quit IRC22:37
mordredzul: because you do not WANT pkg_resources enforcing depends - you're doing that better with packages22:37
zulmordred:  how does that affect the tarball thats already created?22:38
mordredzul: it affects what happens at setup.py install time22:38
zulmordred:  ok ill play around with it then22:38
mordredat install time, it will skip both the pip step, and the inject-requirmeents-into-install-requirements step22:38
*** zaro0508 has joined #openstack-infra22:38
mordredso that not only will pip not attempt to even read the requirements.txt file22:38
mordredbut also, the installed package will never complain at the python level that the version of a thing is wrong22:39
mordredzul: I wrote that feature for you personally22:39
zulmordred:  problem with that if the debian/control is not complete then we might end up with broken packages22:39
zulmordred:  thanks i feel honored22:39
openstackgerritA change was merged to openstack-infra/zuul: Remove usage of d2to1  https://review.openstack.org/4829422:39
openstackgerritA change was merged to openstack-infra/jeepyb: Use python logging isntead of print  https://review.openstack.org/4831322:39
mordredzul: hrm.22:39
mordredzul: so, hrm22:40
dstufftyou probably want to still populate install_requires22:40
*** rnirmal has quit IRC22:40
dstufftjust not install crap22:40
mordreddstufft: yeah - but with versionless stuf22:40
mordreddstufft: so that whatever the version of the os says, that's right22:40
mordredyeah?22:40
fungizul: how are these sorts of bugs normally solved in distros? carrying a patch to python-babel backporting the chinese locale compilation fix? we just can't patch the babel on pypi so unfortunately, fix means later lib22:40
openstackgerritA change was merged to openstack-infra/zuul: Partial sync with OpenStack requirements.  https://review.openstack.org/4829522:41
openstackgerritA change was merged to openstack-infra/jeepyb: Wait for groups to be created before getting uuids  https://review.openstack.org/4831422:41
dstufftmordred: meh, I wouldn't bother, adding a patch for requirements.txt in the debian package shouldn't be that hard22:41
dstufftmordred: by forcing a packager to make a patch, it's a signal that maybe the versions aren't compat22:42
jeblairjog0, Alex_Gaynor: what's the status of the issue referenced in https://review.openstack.org/#/c/48352/1 ?22:42
mordredindeed22:42
dstufftbesides the "patch the source" stuff in the Os packages certainly has way more real world use than SKIP_PIP_INSTALL :)22:43
dstufftmordred: I would provide the option to not reach out and install stuff though22:43
dstufftthough setuptools already has an option for that, so maybe you could just piggyback off that22:44
mordredwell, we can't22:44
mordredbecause we subvert setuptools' installation anyway22:44
mordredbecause we hate easy_install22:44
dstufftmordred: no actually I think you already do this22:44
mordredI believe we do22:44
jog0jeblair: as far as I know still the same, pypy breaks22:44
mordredwe always use distutils.install (--single-version-externally-managed) which skips setuptools trying to do anything22:45
dstufft--single-version-externally-managed is the cannonical way for setuptools to disable dependency installation, so any non openstack project I believe Os packages are likely to already be using them22:45
mordredyup. we enable --single-version-externally-managed at all times22:45
dstufftso I'd assume (maybe i'm wrong?) that they are already using that flag22:45
mordredwell, they may or may not be - it's irrelvant22:45
zulfungi:  usually the patch is isolated and then its backported22:45
mordredWE do a pip install ourselves in pbr22:46
openstackgerritA change was merged to openstack-infra/gear: Fix geard logging config  https://review.openstack.org/4833422:46
mordredwhich is what SKIP_PIP_INSTALL skips22:46
dstufftmordred: but wether it's pip or setuptools doing the install is irrelevant22:46
mordredzul, dstufft: I was wrong, btw, the install_requires still gets populated22:46
mordreddstufft: it's very relevant for us22:46
mordredbut not for this22:46
dstufftmordred: I mean for a packager :)22:46
mordreddstufft: yes. the packagers want it to not happen :)_22:46
dstufftthey either want the install to install deps, or they don't want it22:46
*** thedodd has quit IRC22:47
dstufftin what case would that want easy_install to install dependencies but not pip22:47
dstufftthey*22:47
*** anteaya has quit IRC22:47
mordredyup. so they need to set SKIP_PIP_INSTALL - and honestly, we should probably get that added to dh_python22:47
mordreddstufft: oh - you're saying that we should honor --single-version-externally-managed as way to switch the SKIP_PIP_INSTALL logic?22:47
fungizul: makes sense. but with most of the third-party modules we use not providing stable backport versions for bugs we encounter, the alternative (also not without merit) is to declare that close to a release we'll refuse to fix any bugs which require newer versions of dependencies22:47
dstufftmordred: yes22:48
mordreddstufft: interesting. good idea. I will see if I can do it :)22:48
mordred(setuptools. arg parsgin. evil. crazy. AAAAGGGGHHHH.)22:48
dstufftmordred: at least that would make you act more like a normal python package, just your installs happen with pip instead of setuptools, but unless someone wants to install dependencies specifically with setuptools I don't see a need for a special command to specifically disable pip22:49
zulfungi: right i know we need to find a balance but the way that dpkg/pydist/setuptools drives me batty/causes pain22:49
fungizul: i think it just requires more of a dedication on the upstream side to accept that there are certain bugs which are better left in the release if doing otherwise means more dependency churn for packagers22:49
mordreddstufft: I guess the thing is is that --single-version-externally-managed as a mechanism to disable internet depends seems like relying on an impl hack22:49
mordreddstufft: --single-version-externally-managed is suppoed to turn off egg installs22:50
mordredas a result, it does it by punting to distutils22:50
mordredwhich does not have internet downloading of depends22:50
*** esker has quit IRC22:50
zulfungi:  sure but you are dealing with different distros with different intests different release schedules22:50
mordredso using that flag as a flag to mean "please skip downloading things from the internet" seem VERY FAR from being something that a sane person would be expected to know22:50
mordredeven though it does have that side effect22:50
dstufftmordred: python setup.py install --help says ->22:51
dstufft --single-version-externally-managed  used by system package builders to22:51
dstufft                                       create 'flat' eggs22:51
*** thingee is now known as thingee_zzz22:51
dstufftso it's called out in the help as the thing system package builders should use22:51
dstufftI mean it's name is terrible22:51
fungizul: getting everyone on board with "no new features requiring new(er versions of) dependencies allowed close to release time" was fairly easy. the "no fixing bugs" proposition will be a harder sell22:51
mordredyes. it says nothing about "disables downloading from the internet"22:51
mordredeven though it does that22:52
lifelessdstufft: the help doesn't imply the thing we care about...22:52
zulbbl need to put people to bed22:52
mordredSKIP_PIP_INSTALL is at least clear as to _exactly_ what it actually does22:52
mordred(again, not arguing that I shouldn't also respond to it for ease of use ...)22:52
*** alexpilotti has quit IRC22:52
dstufftmordred: lifeless to make matters worse, python setup.py develop as a -N // --no-deps22:53
dstuffthas a*22:53
mordreddstufft: yah. I tried to use that first22:53
mordreduntil I discovered it was only for develop22:53
mordredjog0: https://review.openstack.org/#/c/48352/22:53
dstufftmordred: I mean I don't care if you keep the SKIP_PIP_INSTALL thing, but I think that it won't be that useful because anyone who doesn't want to install things probably doesn't want to install things for any python package and is already using --single-version-externally-managed22:54
mordreddstufft: I think you'd be surpised22:54
dstufftmaybe i'm wrong and OS packagers are just downloading crap from the internet22:54
mordredthey all complain at me22:54
*** zaro0508 has quit IRC22:54
mordredBUT - possibly it's a knock on and I'm only noticing it because I'm not honoring single-version-externally-managed22:55
dstufftI think you are aren't you?22:55
jog0mordred:  what about https://review.openstack.org/#/c/48352/ ?22:55
dstufftnow that I think about it22:55
dstufftelse you'd do some wierd recursive shit from pip?22:55
mordredI mean, the flag has no effect on our packages22:55
fungisome prefer to build packages in a chroot with no internet access so they can watch if there are failures related to missing dependencies, since pip install will try to retrieve them22:55
mordredwe _always_ don't make eggs, and we _always_ use pip to processes dependencies22:56
dstufftmordred: right, but when you ``pip install foo`` you don't disable pbr's pip stuff then?22:56
mordredsingle-version-externally-managed has zero impact on anything22:56
dstufftwhere foo is a pbr thing22:56
mordrednope22:56
dstufftwat22:56
mordredit works22:56
dstufftdon't do that22:56
mordredbecuse the calling pip does the egg_info22:56
mordredand installs the list22:56
dstufftit still calls python setup.py install22:57
dstufftand uses --single-version-externally-managed to keep it from installing stuff22:57
dstufft(among other things)22:57
mordredHAHAHAHAHAHAHAHAHAAHHAAH22:57
mordredok22:57
mordredseriously22:57
mordredI'll do the flag22:57
mordredbut please tell me you realize that interface is ludicrous and non-documented and magic22:57
mordredand there is NO WAY I could be expected to know that22:57
*** fbo is now known as fbo_away22:58
openstackgerritElizabeth Krumbach Joseph proposed a change to openstack-infra/config: Add initial files for tripleo in the gate.  https://review.openstack.org/4945422:58
pleia2I get a gold star \o/22:58
pleia2(it probably needs work)22:58
lifeless\o/22:58
fungipleia2: the star is foil paper with water-activated adhesive backing. enjoy22:59
mordredjog0: a) why is it failing pypy, b) isn't the point of the pypy gate to prevent things from breaking on pypy?22:59
pleia2fungi: goodie!22:59
dstufftmordred: everything about setuptools is fucking terrible22:59
mordreddstufft: so terrible22:59
dstufftI don't expect anyone to understand how it works22:59
dstuffthell I don't fully understand how it works23:00
mordreddstufft: so seriously? pip --single-version-externally-managed the things that it installs via depends?23:00
dstufftand I spend a not insignicant amount of time working on it23:00
dstufftmordred: yes, sec I'll show you the code23:00
mordredwow23:00
* mordred loves that as much as he thinks he understands about the weird back corners of things, he's always able to learn something new23:00
dstuffthttps://github.com/pypa/pip/blob/develop/pip/req.py#L61123:00
*** anteaya has joined #openstack-infra23:01
jog0mordred: why don't know and Alex_Gaynor didn't either23:02
jog0b) yes, but if we don't know the issue and can't fix it soon (when all RCs are out) don't see why we should hold this up for a possible pypy bug23:02
dstufftmordred: https://github.com/openstack-dev/pbr/blob/master/pbr/packaging.py#L357-L358 ?23:03
*** sarob_ has joined #openstack-infra23:04
dstufftmordred:  a little scatter brained today, I'm sick :(23:04
*** sarob_ has quit IRC23:05
*** sarob_ has joined #openstack-infra23:05
*** sarob has quit IRC23:06
*** dkliban has quit IRC23:08
*** sarob_ has quit IRC23:08
*** AlexF has quit IRC23:14
mordredjog0: ok.23:15
mordreddstufft: yuck! sick seems to be going around this week23:15
clarkbis elastic-recheck bot still derping? I am finally home, eating lunch then will push in gearman worker changes23:16
jog0fungi: I think the issue is gerritlib is dropping things23:19
jog0I am seeing patches pop up on the bot that I don't see23:19
*** boris-42 has quit IRC23:19
jog0hmm actually I see the last patch the bot detected23:20
jog0https://review.openstack.org/#/c/49435/223:20
*** mdenny has quit IRC23:20
jog0fungi: can I get the logs for everything after that23:20
jog0mtreinish: ^23:22
*** rcleere has quit IRC23:22
*** julim has quit IRC23:23
*** dcramer_ has joined #openstack-infra23:30
*** rfolco has joined #openstack-infra23:30
*** giulivo has quit IRC23:31
*** jergerber has quit IRC23:31
anteayaclarkb: glad you are home safe23:34
*** Ryan_Lane has quit IRC23:44
*** gyee has quit IRC23:46
jog0clarkb: it is its workign locally for me (non-bot mode)23:46
jog0clarkb: also status.openstack.org/elastic-recheck/23:48
jog0jeblair did that one23:48
clarkbjog0: nice23:50
clarkbjog0: the irc bot is not reporting on stuff right? I am about to pull out the laptop and start poking at it23:50
jog0clarkb: correct23:50
*** rfolco has quit IRC23:58

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!