Sunday, 2014-06-08

*** sarob_ has quit IRC00:00
*** otherwiseguy has quit IRC00:07
mordredSpamapS: I still stand by my statement that any of these services having ports at all is stupid00:11
*** Ryan_Lane has joined #openstack-infra00:13
SpamapSmordred: keystone is the one...00:13
*** otherwiseguy has joined #openstack-infra00:13
mordredSpamapS: keystone's port should be "443"00:13
SpamapSyeah or even 80 with starttls .. I would agree with that position. But .. Ipv4 strikes again. :-/00:14
mordredSpamapS: you know - I should hit "https://keystone.hpcloud.net" - which should give me other endpoints which could look like "https://glance.hpcloud.net"00:14
mordredOR00:14
mordred"https://hpcloud.net/keystone" and "https://hpcloud.net/glance"00:14
mordredthis has been a solved problem for 30 years00:14
mordredwhy the HELL we're re-inventing it and making it harder is beyond me00:15
*** sarob has joined #openstack-infra00:15
mordredSpamapS: I do not see what IPv4 has to do with anything00:15
mordredunless the problem is "can't figure out how to configure virtualhost and proxypass"00:16
mordredin which case - stop trying to run cloud software and go back to writing iphone apps00:16
* mordred grumps around and strokes his beard00:16
SpamapSmordred: Well, SSL requires a 1 IP to 1 certificate relationship....00:16
mordredSpamapS: yup. but you already need an ip for the https://hpcloud.net entrypoint00:17
mordredif you LB/proxypass through URLS to the sub-endpoints you're fine if you only have 1 IP00:17
mordredof course, if you only have 1 ip and you're tring to run a cloud you're going to run in to other issues ...00:17
mordred:)00:17
SpamapSmordred: you're pushing an architecture on people.. we can't possibly ask people to do things the usual way. EVER.00:18
SpamapSThey'll... they'll cry.00:18
*** gokrokve has joined #openstack-infra00:18
mordredSpamapS: oh. right. I'm sorry. I forgot. we should make terrible decisions because someone somewhere might not want to learn antyhing about anything00:18
mordredand would prefer to just hipster their apple to make bouncing icons and expect that to produce a working cloud00:19
SpamapSmordred: or even worse, someone, somewhere, might have learned something else than the entire rest of the ops world learned.00:19
*** sarob has quit IRC00:19
*** otherwiseguy has quit IRC00:19
mordredSpamapS: btw- matsunobu-san wrote a great blog-post on the new-improved semi-sync replication00:19
mordredSpamapS: it has a new mode00:20
mordredthat's pretty good00:20
mordredSpamapS: which of course made me think about all of the "MySQL can't possibly be used at scale for real things" people :)00:20
mordredSpamapS: http://yoshinorimatsunobu.blogspot.com/2014/04/semi-synchronous-replication-at-facebook.html00:21
mordredother people who want to learn about how massive scale mysql deployments really deal with HA ^^00:21
SpamapSsemi-sync was pretty good already for the DR case IIRC00:22
mordredSpamapS: yeah - they added a new thing called "loss-less semi-sync"00:22
mordredit changes when the master commit is commited, so that you don't potentially show a commit to other clients before it's propagated to at least one slave00:23
mordredSpamapS: which allows them to be able to combine it with slave promotion safely00:23
mordredSpamapS: also, matsunobu-san is pretty ludicrously smart and thorough00:24
SpamapSmordred: nice00:26
SpamapSmordred: I still think Galera is easier to reason about and thus a more generically consumable thing.00:27
SpamapSmordred: at the cost of requiring your app to be resilient to rollbacks.00:28
SpamapSand that whole "have to write everything everywhere" bit :)00:29
mordredyah00:29
mordredI agree - galera seems to be a great small to mid-sized solution00:29
mordredI think we may want to think about a loss-less semi-sync deployment and what that would look like just for completeness - but it may be more trouble than it's worth until we start hitting mega clouds00:30
SpamapSmordred: thing is, I'd almost go for sharding galera before I'd go for the loss-less semi-sync because our challenge isn't page load time... it is simply scale-out00:32
SpamapSbut yeah.. at some point.. when we're topping 500,000 vms and 10k + tenants .. we'll probably want something better00:35
SpamapSlike MongoDB00:35
SpamapScause its webscale00:35
* SpamapS goes back to Saturday00:35
harlowja_at_homerewrite it all in node.js to, webscale -> infinity00:37
*** adrian_otto has quit IRC00:37
*** harlowja_at_home has quit IRC00:57
mordredgod. I look away for a second and I come back and see the two solutinos that have been suggested as "mongo" and "node.js"01:07
*** aconrad_ has joined #openstack-infra01:14
*** sarob has joined #openstack-infra01:15
*** alexpilotti has joined #openstack-infra01:15
*** yamahata has joined #openstack-infra01:19
*** sarob has quit IRC01:19
*** alexpilotti has quit IRC01:24
fungiSpamapS: domain wildcard certs let you do http 1.1 host headers to have multiple https subdomains served as vhosts on one common ip address01:31
fungithough srv rrs in dns could be another option for things like application clients too (unlike for browsers, where they still haven't caught on after more than a decade)01:33
*** Ryan_Lane has quit IRC01:37
*** michchap has joined #openstack-infra01:42
*** rustlebee has quit IRC01:54
*** wenlock has quit IRC01:58
*** homeless has quit IRC02:15
*** sarob has joined #openstack-infra02:15
*** wenlock has joined #openstack-infra02:16
*** homeless has joined #openstack-infra02:16
*** sarob has quit IRC02:19
*** otherwiseguy has joined #openstack-infra02:23
*** praneshp has quit IRC02:32
*** greghaynes has quit IRC02:34
*** otherwiseguy has quit IRC02:36
*** olaph has quit IRC02:52
*** Longgeek has joined #openstack-infra02:52
*** harlowja_at_home has joined #openstack-infra02:53
*** harlowja_at_home has quit IRC02:57
*** mriedem has joined #openstack-infra03:01
*** otherwiseguy has joined #openstack-infra03:03
*** mriedem has quit IRC03:07
*** penguinRaider has quit IRC03:08
*** changbl has quit IRC03:08
*** otherwiseguy has quit IRC03:11
*** sarob has joined #openstack-infra03:15
*** sarob has quit IRC03:19
*** otherwiseguy has joined #openstack-infra03:23
*** harlowja_at_home has joined #openstack-infra03:23
*** fifieldt_ has quit IRC03:23
*** fifieldt_ has joined #openstack-infra03:24
*** Ryan_Lane has joined #openstack-infra03:39
*** harlowja_at_home has quit IRC03:41
*** penguinRaider has joined #openstack-infra03:48
*** otherwiseguy has quit IRC03:48
*** Longgeek has quit IRC03:52
*** Longgeek has joined #openstack-infra04:00
*** Ryan_Lane has quit IRC04:00
*** homeless has quit IRC04:10
*** homeless has joined #openstack-infra04:10
*** godp1301 has joined #openstack-infra04:12
*** sarob has joined #openstack-infra04:15
*** godp1301 has quit IRC04:15
*** sarob has quit IRC04:19
*** Longgeek has quit IRC04:27
*** harlowja_at_home has joined #openstack-infra04:41
*** Longgeek has joined #openstack-infra04:43
*** yfried has joined #openstack-infra05:00
*** talluri has joined #openstack-infra05:03
*** sarob has joined #openstack-infra05:15
*** sarob has quit IRC05:20
*** Longgeek has quit IRC05:25
*** hunner has joined #openstack-infra05:31
*** hunner has joined #openstack-infra05:31
*** hunner is now known as Hunner05:32
*** harlowja_at_home has quit IRC05:53
*** basha has joined #openstack-infra06:14
*** sarob has joined #openstack-infra06:15
*** wenlock has quit IRC06:17
*** sarob has quit IRC06:19
*** Longgeek has joined #openstack-infra06:20
*** praneshp has joined #openstack-infra06:26
*** Longgeek has quit IRC06:29
*** Longgeek has joined #openstack-infra06:29
*** homeless has quit IRC06:38
*** talluri has quit IRC07:04
*** nati_ueno has joined #openstack-infra07:04
*** talluri has joined #openstack-infra07:04
*** pcrews has quit IRC07:05
*** talluri has quit IRC07:09
*** sarob has joined #openstack-infra07:10
*** sarob has quit IRC07:14
*** sarob has joined #openstack-infra07:15
*** sarob has quit IRC07:19
*** basha has quit IRC07:22
*** rcarrill` has joined #openstack-infra07:31
*** rcarrillocruz has quit IRC07:32
*** talluri has joined #openstack-infra07:35
*** talluri has quit IRC07:39
*** basha has joined #openstack-infra07:40
*** _nadya_ has joined #openstack-infra07:41
*** _nadya_ has quit IRC07:42
*** kmartin has quit IRC07:44
*** kmartin has joined #openstack-infra07:44
*** basha has quit IRC07:45
*** kmartin has quit IRC07:47
*** kmartin has joined #openstack-infra07:47
*** basha has joined #openstack-infra07:55
*** basha has quit IRC07:57
*** _nadya_ has joined #openstack-infra08:03
*** talluri has joined #openstack-infra08:04
*** praneshp has quit IRC08:07
*** talluri has quit IRC08:08
*** sarob has joined #openstack-infra08:15
*** basha has joined #openstack-infra08:20
*** sarob has quit IRC08:20
*** mestery has quit IRC08:29
*** mestery_ has joined #openstack-infra08:29
*** nati_ueno has quit IRC08:31
*** andreykurilin_ has joined #openstack-infra08:34
*** zzelle has joined #openstack-infra08:45
*** _nadya_ has quit IRC08:45
*** basha has quit IRC08:47
*** achuprin has quit IRC08:59
*** talluri has joined #openstack-infra09:04
*** talluri has quit IRC09:09
*** achuprin has joined #openstack-infra09:14
*** sarob has joined #openstack-infra09:15
*** toabctl has joined #openstack-infra09:16
*** sarob has quit IRC09:20
*** basha has joined #openstack-infra09:40
*** andreykurilin_ has quit IRC10:09
*** alexpilotti has joined #openstack-infra10:10
*** sarob has joined #openstack-infra10:15
*** sarob has quit IRC10:19
*** basha has quit IRC10:20
*** alexpilotti has quit IRC10:21
*** basha has joined #openstack-infra10:26
*** HenryG has quit IRC10:26
*** basha has quit IRC10:55
*** salv-orlando has quit IRC11:01
*** rcarrill` has quit IRC11:02
*** sarob has joined #openstack-infra11:15
*** sarob has quit IRC11:19
*** lcheng has joined #openstack-infra11:45
*** paul-- has quit IRC12:00
*** paul-- has joined #openstack-infra12:02
*** tnurlygayanov has quit IRC12:03
*** tnurlygayanov has joined #openstack-infra12:06
*** sarob has joined #openstack-infra12:15
*** sarob has quit IRC12:20
*** jerryz has quit IRC12:20
*** yolanda has quit IRC12:21
*** mriedem has joined #openstack-infra12:23
*** lcheng has quit IRC12:33
*** trinaths has joined #openstack-infra12:43
trinathsHello all12:43
openstackgerritMatt Riedemann proposed a change to openstack-infra/elastic-recheck: Make query for ssh timeout bug 1298472 more general  https://review.openstack.org/9863112:44
uvirtbotLaunchpad bug 1298472 in tempest "tempest.scenario.test_volume_boot_pattern.TestVolumeBootPattern.test_volume_boot_pattern fails with TimeoutException on SSH connection" [Undecided,In progress] https://launchpad.net/bugs/129847212:44
*** adalbas has joined #openstack-infra13:05
*** sarob has joined #openstack-infra13:15
*** sarob has quit IRC13:20
*** plars is now known as plars-away13:43
*** zul has quit IRC13:43
*** zul has joined #openstack-infra13:47
*** roeyc has joined #openstack-infra13:54
*** changbl has joined #openstack-infra14:01
*** adrian_otto has joined #openstack-infra14:04
*** adrian_otto has quit IRC14:09
*** adrian_otto has joined #openstack-infra14:09
*** adrian_otto has quit IRC14:14
*** sarob has joined #openstack-infra14:15
*** sarob has quit IRC14:19
*** HenryG has joined #openstack-infra14:26
*** HenryG has quit IRC14:27
*** jhesketh has quit IRC14:39
*** ramashri has joined #openstack-infra14:43
openstackgerritA change was merged to openstack-infra/elastic-recheck: Make query for ssh timeout bug 1298472 more general  https://review.openstack.org/9863114:46
uvirtbotLaunchpad bug 1298472 in tempest "SSHTimeout in tempest scenario tests using nova-network" [Undecided,In progress] https://launchpad.net/bugs/129847214:46
*** lbragstad has quit IRC14:48
*** timrc-afk is now known as timrc14:48
*** roeyc has quit IRC14:50
*** jhesketh has joined #openstack-infra14:51
*** homeless has joined #openstack-infra15:00
*** gokrokve has quit IRC15:04
*** gokrokve has joined #openstack-infra15:04
*** jhesketh has quit IRC15:05
*** sarob has joined #openstack-infra15:15
*** wenlock has joined #openstack-infra15:15
*** portante has quit IRC15:16
*** gema has quit IRC15:16
*** flashgordon has quit IRC15:16
*** flashgordon has joined #openstack-infra15:16
*** gema has joined #openstack-infra15:16
*** portante has joined #openstack-infra15:16
*** jhesketh has joined #openstack-infra15:18
*** talluri has joined #openstack-infra15:20
*** sarob has quit IRC15:20
SpamapSfungi: true the wildcard cert does do that, and most if not all API-driving libraries will support wildcards now.15:20
*** flaper87|afk has quit IRC15:23
*** flaper87|afk has joined #openstack-infra15:23
fungiSpamapS: subject alt names in a cert are also a possibility there for supporting multiple domains off one cert15:24
fungithough again, client-side support has been spotty in the past (but may be mostly a non-issue with modern libraries)15:25
SpamapSfungi: Last I checked (2010 time frame) that RFC was contentious.15:34
SpamapSperhaps it was put to bed15:34
*** HenryG has joined #openstack-infra15:37
*** portante has quit IRC15:42
*** portante has joined #openstack-infra15:42
*** HenryG has quit IRC15:42
*** HenryG has joined #openstack-infra15:42
*** yfried has quit IRC15:42
*** ramashri has quit IRC15:44
mordredSergeyLukjanov: you aroudn?15:51
*** homeless_ has joined #openstack-infra15:52
* mordred still contends that even if you need to support one-ip-per-cert that the ports don't really solve anything15:54
*** homeless has quit IRC15:54
mordredbecause you can _totally_ choose to install an apache or nginx with a rewrite/proxypass in front of a bunch of services running on non-public ports to map services to path locations15:55
mordredso if you're short on ipv4, there is a way to deal with it that doesn't involve registering iana ports15:55
*** wenlock has quit IRC15:59
*** wenlock has joined #openstack-infra16:06
*** homeless_ has quit IRC16:11
*** adalbas has quit IRC16:12
*** gokrokve has quit IRC16:15
*** sarob has joined #openstack-infra16:15
*** gokrokve has joined #openstack-infra16:15
*** aconrad_ has quit IRC16:16
*** sarob has quit IRC16:19
*** gokrokve has quit IRC16:20
*** timrc is now known as timrc-afk16:20
*** andreykurilin_ has joined #openstack-infra16:22
*** penguinRaider has quit IRC16:23
*** homeless has joined #openstack-infra16:24
*** talluri has quit IRC16:27
openstackgerritAlex Gaynor proposed a change to openstack-infra/config: Added a puppet module for livegrep.  https://review.openstack.org/9687716:35
*** yamahata has quit IRC16:51
*** harlowja_at_home has joined #openstack-infra16:52
*** adrian_otto has joined #openstack-infra17:04
anteayasaw a cow and a calf moose and stared with a wolf this weekend, missed seeing the bear17:06
anteayaso all in all, pretty good17:06
*** rcarrillocruz has joined #openstack-infra17:06
*** gokrokve has joined #openstack-infra17:06
*** talluri has joined #openstack-infra17:07
mordredanteaya: excellent17:09
anteayamordred: any wildlife in your weekend, so far?17:10
anteayaor just you?17:10
*** gokrokve_ has joined #openstack-infra17:11
*** gokrokve has quit IRC17:11
*** tkelsey has joined #openstack-infra17:13
*** sarob has joined #openstack-infra17:15
*** godp1301 has joined #openstack-infra17:16
openstackgerritChristian Berendt proposed a change to openstack-infra/bugdaystats: Prettified all HTML files  https://review.openstack.org/9865317:16
trinathsanteaya: you went to a zoo today.. ?17:16
anteayano17:17
anteayaI walked around outside17:17
anteayathere are moose and wolves and bear where I live17:18
openstackgerritChristian Berendt proposed a change to openstack-infra/publications: Prettified all HTML files  https://review.openstack.org/9865417:18
trinathsanteaya: oh!. okay.. living very near to wild life.17:18
anteayayeah17:19
anteayaand trees and water17:19
*** sarob has quit IRC17:19
*** gokrokve_ has quit IRC17:20
trinathsanteaya: okay. surrounded by green17:20
anteayayeah17:20
*** gokrokve has joined #openstack-infra17:20
trinathsdaily you go for a walk.. hearing music.. of birds and air..17:21
*** tkelsey has quit IRC17:21
openstackgerritChristian Berendt proposed a change to openstack-infra/reviewday: Prettified all HTML files  https://review.openstack.org/9865517:21
*** harlowja_at_home has quit IRC17:22
openstackgerritMonty Taylor proposed a change to openstack-infra/config: Run stackalytics in infra  https://review.openstack.org/9865617:22
mordredSergeyLukjanov: ^^ there is one thing about that that I'm unsure of - namely when and how stackalytics-processor should be run - but I think that's good enough to start17:23
mordredSergeyLukjanov: I'm going to file a few bugs/feature requests after having dived in a little bit - what's the best way to chat with Ilya - is he aroudn on IRC?17:23
*** jpeeler has quit IRC17:23
*** Shrews has quit IRC17:23
*** arunkant has quit IRC17:23
*** cyli has quit IRC17:23
*** mdenny has quit IRC17:23
*** isviridov|away has quit IRC17:23
anteayayep17:23
*** cyli has joined #openstack-infra17:24
mordredawesome17:24
*** isviridov has joined #openstack-infra17:24
*** mdenny has joined #openstack-infra17:24
*** jpeeler has joined #openstack-infra17:24
*** jpeeler has joined #openstack-infra17:24
*** arunkant has joined #openstack-infra17:24
anteayaare we bring stackalytics under infra?17:24
mordredoh. yeah. there he is17:24
mordredanteaya: yeah - we chatted about it briefly at the summit17:24
anteayaI ether missed that or I missed that17:24
anteayabut yay17:24
mordredwell, it was more informal ish - there may have been beer17:25
anteayaah17:25
anteayaI don't think I saw you at night anywhere17:25
*** Shrews has joined #openstack-infra17:25
mordredbut in general, it's become a thing that people look at a lot, so it seemed that we should bring it in17:25
*** gokrokve has quit IRC17:25
mordredanteaya: I agree with that17:25
mordredanteaya: I feel like I saw you for like 15 minutes17:25
anteayaI saw you for about 2 minutes during daytime, I remember that17:25
anteayaby the waterfall17:25
anteayaah okay17:25
anteayatime moves more slowly for me apparently17:26
mordredilyashakhat: what sorts of ways would you prefer for me to make feature requests?17:26
fungii think it also bears mentioning in a reply to http://lists.openstack.org/pipermail/openstack-infra/2014-June/001292.html17:26
mordredfungi: yah. funny story - I was in the middle of writing a response to that17:26
*** godp1301 has quit IRC17:26
mordredfungi: I was going to point them to that patch as an example of a patch they could cargo cult17:27
pleia2good, once we bring that in I'd like to add the reviewstats project to that server too17:27
fungiperfect17:27
* fungi fades back into the weekend17:27
* pleia2 weekend too17:27
*** wenlock has quit IRC17:27
mordredyah. there's a thing the activity-board email mentions that may be worth thinking about17:28
fungii have an entire server room here i need to disassemble and junk. still need to move some of it into a virtual server provider first. should get back to it :/17:28
mordredwhich is the idea of having _one_ thing do the data collection, and then have various thigns be able to grab that data17:28
mordredso if we have reviewstats, and stackalytics and activity-board, then perhaps at some point in the future a single "I know how to collect data" thing can emerge17:29
fungiyes, having a raw data earehouse for the various stats-generating tools to mine would be doubleplusgood17:29
fungis/earhouse/warehouse/17:29
mordredright now, I'd actually imagine that's likely to be stackalytics since it does git, gerrit and launchpad and has an API17:29
funginot sure what an earhouse is17:29
mordredbut I could be wrote17:29
mordredfungi: if you don't know what an earhouse is, then I can't explain it17:30
fungiin fact, i cannot explain it, no17:30
fungian earhouse is where you keep your ear wigs17:30
*** trinaths has quit IRC17:31
*** pcrews has joined #openstack-infra17:32
*** ildikov_ has joined #openstack-infra17:34
openstackgerritMonty Taylor proposed a change to openstack-infra/config: Run stackalytics in infra  https://review.openstack.org/9865617:35
*** Alec-A2 has joined #openstack-infra17:37
*** ildikov has quit IRC17:38
anteayaI don't want an earhouse17:39
mordredfungi: btw - I've spun up a server, added a key to hiera and am running the puppet above ^^ on it by hand out of the mordred environment17:40
*** yfried has joined #openstack-infra17:40
*** yfried has joined #openstack-infra17:40
openstackgerritMonty Taylor proposed a change to openstack-infra/config: Run stackalytics in infra  https://review.openstack.org/9865617:41
*** gema has quit IRC17:44
*** gema has joined #openstack-infra17:46
*** gokrokve has joined #openstack-infra17:54
openstackgerritMonty Taylor proposed a change to openstack-infra/config: Run stackalytics in infra  https://review.openstack.org/9865617:59
*** ildikov_ has quit IRC18:01
*** ildikov_ has joined #openstack-infra18:02
mordredilyashakhat: _retrieve_project_list_from_github should grab list-projects from gerrit and not github :)18:02
YorikSarfungi: On stackalytics-processor: I've heard it's run once in 4 hours by cron now.18:03
mordredYorikSar: cool. I was sort of imagining something like that might be the case - it's not in the "HowToRun" doc though18:03
YorikSarfungi, mordred: and bout bugs and features. It's on stackforge - I guess noone would object using https://bugs.launchpad.net/stackalytics18:04
mordredYorikSar: awesome! I will file bugs18:05
*** andreykurilin_ has quit IRC18:05
* YorikSar hopes ilyashakhat won't bite in the Monday morning18:06
YorikSarmordred: Oh... You're going to run it in Apache...18:08
mestery_fungi: FYI, I don't think the failure in the VM you saved is the same one. armax's code should have kept the guest around, and it didn't.18:08
mestery_fungi: So you can recycle that one please. If it triggers again, we can have anther look. I'll see about recreating it locally tomorrow.18:08
mestery_fungi: Have to head out to a family gathering for hte rest of the day. Thanks and enjoy your Sunday!18:09
*** mestery_ is now known as mestery18:09
YorikSarmordred: I know that stackalytics web process is very memory-hungry. It holds lots of data and indexes in memory. I'm not sure as of how Apache would react o that.18:09
mordredYorikSar: well, we'll have to poke around at it - if we need to do an out-of-process and have apache do proxypass that's possible too18:10
mordredYorikSar: we just try to keep things as consistent across the board and don't have nginx anywhere18:10
YorikSarmordred: Currently they're using uwsgi there.18:11
YorikSarmordred: I guess there's no dependency on Nginx - Apache can proxy to uwsgi as well.18:12
*** sarob has joined #openstack-infra18:15
mordredYorikSar: from apt-cache search it seems you have to install nginx to get uwsgi installed18:18
YorikSarmordred: Weird Ubuntu...18:18
mordredYorikSar: mordred@camelot:~/src/stackforge/stackalytics$ apt-cache search uwsgi18:19
mordrednginx-extras - nginx web/proxy server (extended version)18:19
mordrednginx-full - nginx web/proxy server (standard version)18:19
mordred:(18:19
*** sarob has quit IRC18:20
mordredYorikSar: oh - pip install uwsgi18:20
mordredYorikSar: k. nevermind. I can work with uwsgi18:20
mordredI'll try that next18:20
YorikSarmordred: On Gentoo uwsgi may depend on Apache :D18:21
YorikSarmordred: I guess it's up to package maintainer's preference :)18:22
mordredlovely18:23
mordredYorikSar: stackalytics doesn't seem to be reading /etc/stackalytics/stackalytics.conf by default18:24
mordredI had to run stackalytics-processor with --config-dir=/etc/stackalytics18:24
*** yfried has quit IRC18:24
mordred(I'm just complaining now)18:24
*** yfried has joined #openstack-infra18:24
YorikSarmordred: That's strange. They use oslo.config...18:26
mordredyeah. I'm wondering if it's not getting initialized properly through a code path ...18:27
*** Longgeek has quit IRC18:30
*** homeless has quit IRC18:30
YorikSarmordred: Aha. project name is not passed to CONF() call.18:31
mordredwell, that's poopy18:32
* mordred makes patch18:32
*** pcrews has quit IRC18:32
* mordred is currently making a patch to remove github urls18:32
pleia2\o/18:32
*** pcrews has joined #openstack-infra18:34
* YorikSar looks how mordred is going to become active contributor to stackalytics18:34
mordredYorikSar: I don't even see the CONF() call18:35
mordredah. I see it now18:36
YorikSarhttps://github.com/stackforge/stackalytics/blob/master/stackalytics/processor/main.py#L26618:36
mordredyah. got that18:36
YorikSarmordred: I'm too slow this evening18:37
mordrednow, in the web app, it looks like it's expecting me to pass in an env var18:37
YorikSarmordred: Yeah...18:38
mordredYorikSar: https://review.openstack.org/9865918:40
YorikSarmordred: You shoulld pass project='stackalytics' to the call or default locations won't work18:41
YorikSarmordred: Oh, wait... Is project the first argument?18:41
mordredYorikSar: think so - that's how we do it in storyboard18:41
mordredYorikSar: also, is this: logging.setup('stackalytics.dashboard')18:42
YorikSarhttps://github.com/openstack/oslo.config/blob/master/oslo/config/cfg.py#L159618:42
mordredYorikSar: valid usage? it is, right?18:42
mordredoh. piddle18:42
YorikSarmordred: It looks it's the second. Anyway, cryptic positional args should be avoided.18:42
*** praneshp has joined #openstack-infra18:43
mordredYorikSar: yah. just pushed an update18:44
YorikSarmordred: Oh, here it is: https://pypi.python.org/pypi/kwonly18:45
YorikSarmordred: I guess, some libs should enforce this more often.18:46
YorikSarmordred: Although I'm barely involved in stackalytics here's a +1 from me :)18:51
mordred:)18:51
mordredlifeless: awake?18:53
mordredlifeless: run.py: error: unrecognized arguments: discover -t ./ ./tests --list18:57
* anteaya readies to move locations19:02
*** yfried has quit IRC19:06
*** yfried has joined #openstack-infra19:06
*** gokrokve has quit IRC19:07
*** gokrokve has joined #openstack-infra19:07
openstackgerritMonty Taylor proposed a change to openstack-infra/config: Run stackalytics in infra  https://review.openstack.org/9865619:08
Alex_Gaynormordred: ping19:10
*** gokrokve has quit IRC19:11
mordredAlex_Gaynor: heyho19:12
mordredYorikSar: ok. I fixed a thing and also uploaded another patch to move dashboard into stackalytics :)19:15
*** sarob has joined #openstack-infra19:15
Alex_Gaynormordred: So, my understanding is that the desire for config stuff is that the modules are standalone, and then the manifest is where the openstack specific stuff. Right now for livegrep it's just sort of all in one place, what's the right way to split that stuff up?19:16
*** talluri has quit IRC19:16
mordredAlex_Gaynor: so - there should be a module that's standalone - and then there shoudl be something in openstack_project that includes that module and also insludes openstack_project::server (which puts our base things on it)19:17
YorikSarmordred: Nice. I've been looking for the web-related code there for a long time when I first took a look at it.19:17
*** wenlock has joined #openstack-infra19:18
mordredAlex_Gaynor: and then manifests/site.pp should instantiate the thing in openstack_project19:18
mordredAlex_Gaynor: let me look at your patch real quick19:18
Alex_Gaynorthanks19:19
mordredAlex_Gaynor: that looks fine to me - I may want to go in and make a follow up patch to have it consume projects.yaml instead of keeping its own list - but this is probably great for now19:19
*** sarob has quit IRC19:20
Alex_Gaynormordred: ok. We'd also discussed having it either be a mirror for cgit, or run on the cgit machine itself. Should that be a blocker for getting this landed?19:20
mordredI dont' think so19:20
mordredI think we'll learn something about it from landing this19:20
*** cody-somerville has quit IRC19:20
Alex_GaynorOk, let me remove the notes from teh commit message and put these into a comment in the code19:20
mordredAlex_Gaynor: how does the indexer know to re-index things?19:21
Alex_Gaynormordred: the codesearch service (the indexer) is subscribed to the repos its indexing19:21
*** pcrews has quit IRC19:22
mordredAlex_Gaynor: ah. I see it19:22
openstackgerritAlex Gaynor proposed a change to openstack-infra/config: Added a puppet module for livegrep.  https://review.openstack.org/9687719:23
mordredAlex_Gaynor: I'm going to eat some brunch - but I'll review that a little bit deeper in just a bit19:24
Alex_Gaynormordred: thanks! I'll be around all afternoon19:24
mordredAlex_Gaynor: I, for one, look forward to our new livegrep overlords19:24
Alex_Gaynor:-)19:24
*** wenlock has quit IRC19:29
*** rcarrillocruz has quit IRC19:46
*** rcarrillocruz has joined #openstack-infra19:50
mordredYorikSar: http://stackalytics.openstack.org/20:00
mordredYorikSar: at least works to some degree20:00
mordredfungi: ^^ that's using the puppet module20:00
openstackgerritMonty Taylor proposed a change to openstack-infra/config: Run stackalytics in infra  https://review.openstack.org/9865620:01
mordredwith those changes20:01
mordredoutstanding questions: a) uwsgi b) cron job for processor20:01
fungimordred: darned awesome!20:01
YorikSarHm... It doesn't data in dropdowns...20:01
lifelessmordred: hi20:02
mordredYorikSar: it's still processing data20:02
mordredlifeless: nm. I foudn the error20:02
mordredlifeless: I was having trouble diagnosing a testr issue20:02
lifelessmordred: thats being thrown by a unit test that is consulting the real sys.argv20:02
YorikSarmordred: Ah... Ok.20:02
mordredfungi: I also submitted like 10 patches to stackalytics that I think we should probably get landed before rolling that live20:02
mordredYorikSar: I'm also seeing things like this:20:02
mordred2014-06-08 20:02:50.996 10166 WARNING stackalytics.processor.utils [-] Error while reading uri: HTTP Error 404: Not Found20:02
mordredYorikSar: which I just submitted a patch to add the url to the message :)20:03
fungimordred: not at all unexpected20:03
mordredfungi: mostly little things - the normal result of running one in the second location20:03
mordredfungi: so things that were assumptions about their install that are not the same for ours and not properly picking up config20:04
mordredfungi: but I've got to say in all fairness, it took me relatively low amount of effort to get that up and running and fully puppetted20:04
openstackgerritMonty Taylor proposed a change to openstack-infra/config: Run stackalytics in infra  https://review.openstack.org/9865620:08
mordredYorikSar, fungi: ^^ ok. there it is running processor in cron instead20:08
*** sarob has joined #openstack-infra20:15
*** sarob has quit IRC20:19
*** michchap has quit IRC20:20
*** _nadya_ has joined #openstack-infra20:20
*** _nadya_ has joined #openstack-infra20:28
*** _nadya_ has quit IRC20:30
openstackgerritMonty Taylor proposed a change to openstack-infra/config: Split the database grants into their own class  https://review.openstack.org/9858620:35
openstackgerritMonty Taylor proposed a change to openstack-infra/config: Add support for disk-image-builder in nodepool  https://review.openstack.org/8847920:35
*** praneshp has quit IRC20:37
*** ayoung has quit IRC20:45
*** salv-orlando has joined #openstack-infra20:57
*** zz_gondoi is now known as gondoi20:57
*** adrian_otto has quit IRC21:00
*** ayoung has joined #openstack-infra21:00
*** chuckC has quit IRC21:01
fungipleia2: this just in... DebConf14 - Call for talks, BoFs and events https://lists.debian.org/debian-devel-announce/2014/06/msg00000.html21:13
*** sarob has joined #openstack-infra21:15
*** salv-orlando has quit IRC21:15
lifelesshmm21:19
*** sarob has quit IRC21:19
*** salv-orlando has joined #openstack-infra21:20
*** gondoi is now known as zz_gondoi21:22
*** otter768 has joined #openstack-infra21:24
*** mgagne has quit IRC21:24
fungipleia2: given your posited interest in presenting on possible debian/openstack infra cross-pollination, it might be worth bouncing ideas off Mithrandir (since he's on the dsa team)21:30
*** otter768 has quit IRC21:37
*** otter768 has joined #openstack-infra21:38
*** mgagne has joined #openstack-infra21:40
*** mgagne is now known as Guest8923421:40
*** pcrews has joined #openstack-infra21:40
*** Guest89234 is now known as mgagne121:41
*** mgagne1 has quit IRC21:42
*** mgagne1 has joined #openstack-infra21:42
mgagne1XenServer CI keeps failing on my change, who or which channel should I ask to help me debug? https://review.openstack.org/#/c/91727/21:43
mgagne1All I managed to find is "'amqplib' is not in global-requirements.txt", I'm not sure if it's an error or not. And '/usr/sbin/apache2ctl: command not found' but fungi looks to say it's not a big deal in bug irc://grim.calavera.ca:6697/#132620921:43
mgagne1bug #1326209 *21:44
uvirtbotLaunchpad bug 1326209 in devstack "/usr/sbin/apache2ctl: command not found" [Undecided,New] https://launchpad.net/bugs/132620921:44
fungimgagne1: "XenServer CI" is an unofficial, third-party test system. BobBall may be able to look into it and provide suggestions (i think he may have some insight into it)21:48
fungimgagne1: er, not BobBall (he's citrix)21:49
fungimgagne1: oh, maybe so. the contact address for that account is at citrix.com21:50
* fungi can't keep all the third-party systems straight any longer21:51
mgagne1fungi: right, as a change owner, where can I find information about 3rd party contact? Emails are ok but irc nicknames would be great so I can bug them21:51
*** salv-orlando has quit IRC21:51
mgagne1fungi: and the UX of a change owner about how to debug such results isn't great. No README or explanation on how to analyse the result. :-/21:52
fungimgagne1: there is work in progress to get wiki pages set up for all of them and actively maintained (anteaya is spear-heading that effort). in the meantime, hovering your mouse cursor over the name of the account should give you a valid contact e-mail address21:52
fungimgagne1: and yeah, they need to start including debugging instructions and links to additional details in their respective failure messages21:53
*** pcrews has quit IRC21:55
*** salv-orlando has joined #openstack-infra21:57
*** alkari has joined #openstack-infra22:02
*** alkari1 has joined #openstack-infra22:06
alkari1hello22:07
mordredfungi: actually - you know - instaed of getting debugging instrauctions in the failure message22:10
*** alkari has quit IRC22:10
mordredfungi: if we expand on that "hide all the CI comments" UI patch - we could potentially store a single copy of debug instructions per-test system22:11
mordredfungi: and have the UI thing provide a way to request the info in a discoverable way22:11
openstackgerritMonty Taylor proposed a change to openstack-infra/publications: Use reveal.js instead of slidy  https://review.openstack.org/9867022:14
mgagne1mordred: are the tests ran across all 3rd party the same with the same captured logs?22:14
*** salv-orlando has quit IRC22:15
mordredmgagne1: nope22:15
mordredmgagne1: the tests for a given 3rd party are completely up to the 3rd party22:15
mgagne1mordred: How can debug instructions be the same for all?22:15
*** sarob has joined #openstack-infra22:15
mgagne1mordred: oh, "per test system" right22:15
mordredyah22:15
mgagne1mordred: what if the test system is updated, would having a single copy of the instruction be confusing for someone debugging results from a previous version?22:17
*** otter768 has quit IRC22:18
mordredmgagne1: possibly ... I suppose maybe just having the instructions in the results are maybe not terrible22:18
*** sarob has quit IRC22:20
*** harlowja_at_home has joined #openstack-infra22:21
*** lcheng has joined #openstack-infra22:23
*** aconrad_ has joined #openstack-infra22:23
alkari1folks, im looking for a way to consolidate openstack logs.. any advice here?22:24
*** gokrokve has joined #openstack-infra22:25
*** harlowja_at_home has quit IRC22:32
*** Ryan_Lane has joined #openstack-infra22:34
*** salv-orlando has joined #openstack-infra22:34
*** gokrokve has quit IRC22:36
*** praneshp has joined #openstack-infra22:36
*** gokrokve has joined #openstack-infra22:37
*** adrian_otto has joined #openstack-infra22:38
*** andreykurilin_ has joined #openstack-infra22:38
*** lcheng has quit IRC22:40
*** gokrokve has quit IRC22:41
*** salv-orlando has quit IRC22:42
*** praneshp_ has joined #openstack-infra22:44
mordredalkari1: hi! so, we're mainly the ones who run the openstack developer infrastructure - but we are fans of logstash/elasticsearch22:44
*** praneshp has quit IRC22:46
*** praneshp_ is now known as praneshp22:46
*** salv-orlando has joined #openstack-infra22:46
*** andreykurilin_ has quit IRC22:56
*** pcrews has joined #openstack-infra22:57
*** salv-orlando has quit IRC23:04
*** zzelle has quit IRC23:10
*** sarob has joined #openstack-infra23:15
anteayaXenServer CI is citrix and BobBall is the contact23:15
anteayamgagne1:23:16
* anteaya is looking forward to the renaming23:17
anteaya24 in the gate an only two left at over 40 hours23:18
anteayayay for the weekend23:18
*** sarob has quit IRC23:19
Alex_Gaynoranteaya: pacifica is starting to wake up and get to work I think, hopefully they don't destroy our shot at emptying the queue23:25
anteayapacifica is a gentle sort23:25
anteayahopefully they will be kind23:25
openstackgerritMonty Taylor proposed a change to openstack-infra/publications: Use reveal.js instead of slidy  https://review.openstack.org/9867023:28
mgagne1anteaya: I sent him an email =)23:30
anteayamgagne1: cool23:30
*** _nadya_ has joined #openstack-infra23:30
mgagne1anteaya: yha, both of my changes failed for unclear reason or transient bugs :-/23:31
anteayamgagne1: rather opaque23:32
anteayaBobBall is rather responsive and he also is a contributor23:32
anteayaso he should understand what you are talking about, which is a huge bonus23:32
ianwanteaya: public holiday in Australia today, so that might slow things down23:32
mgagne1anteaya: I'll wait for his answer on Monday.23:32
anteayaianw: ack23:32
anteayaand good, more weekend to clear the queue23:33
anteayamgagne1: very good23:33
mgagne1anteaya: looking forward for those improvements in the 3rd party area =)23:33
anteayame too23:34
anteayait is slow, like pushing mud with a rope, but there is an agreed upon direction23:34
*** _nadya_ has quit IRC23:34
anteayaat least agreed upon within infra, which is a start23:35
mordredanteaya: wow. pushing mud with a rope seems like about the slowest thing possible23:35
anteayaoh no23:36
anteayamuch slower movement23:36
anteayaglaciers are slower23:36
anteayaso are the crystalization of gems inside of cooling rock23:36
anteayathe thing with using rope to push mud is the mud keeps being not in front of the rope23:37
anteayaso you have to pull back and try again23:37
mordrednod23:38
anteayathat's my best analogy atm23:40
*** oomichi has quit IRC23:42
*** yamahata has joined #openstack-infra23:42
*** markmcclain has joined #openstack-infra23:44
*** oomichi has joined #openstack-infra23:44
*** markmcclain1 has joined #openstack-infra23:45
*** markmcclain has quit IRC23:48
anteayamordred: I am going of the tc meeting notes to do the thing I have an action item for23:48
anteayamordred: you too have an action item: ACTION: mordred still to prepare draft binary proposal for the "TC direction" column scores23:49
anteayas/going of/going over23:50
mordredanteaya: thanks23:50
anteayanp23:52
anteayafungi: you had responded to me thusly during the tc meeting: <fungi> anteaya: i can imagine a two-dimensional plot of voter turnout vs. commits23:52
anteayais it realistic to expect that any such plot might see the light of day this week?23:53
*** nati_ueno has joined #openstack-infra23:53
anteayaor how is your expected workload shaping up for the next few days23:53
*** godp1301 has joined #openstack-infra23:55
Alex_GaynorPretty sure the requirements gate isn't going to pass at all right now because of: http://bpaste.net/show/TGgG5c6kY9qTjzKiqqwy/23:58
anteayaAlex_Gaynor: is that an error in six?23:59
Alex_Gaynoranteaya: it looks like it only happens during the installation process, I can't reproduce it except inside of pip. (six did a new release today, that's definitely the cause)23:59
anteayaoh no23:59
anteayano no no23:59
*** godp1301 has quit IRC23:59
mordredAlex_Gaynor: sigh23:59

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