| *** openstack has joined #openstack-infra | 13:49 | |
| *** vhoward- has joined #openstack-infra | 13:51 | |
| *** ociuhandu has joined #openstack-infra | 13:51 | |
| fungi | we lost the openstackgerrit bot at 11:43:28 utc today to a ping timeout | 13:52 |
|---|---|---|
| fungi | i can't find any record of when the openstack meetbot dropped out though, so i must have been netsplit or something when it happened | 13:52 |
| fungi | probably around when my client reconnected yesterday at 00:33 utc | 13:53 |
| *** pcrews has joined #openstack-infra | 13:55 | |
| *** om has quit IRC | 13:56 | |
| *** craigbr has quit IRC | 13:57 | |
| *** rgergan__ has quit IRC | 13:59 | |
| *** CaptTofu has joined #openstack-infra | 13:59 | |
| *** abhishek__ has quit IRC | 14:05 | |
| *** oomichi has quit IRC | 14:08 | |
| *** dkliban_afk is now known as dkliban | 14:09 | |
| *** otherwiseguy has quit IRC | 14:10 | |
| *** markwash has quit IRC | 14:12 | |
| *** sabeen has quit IRC | 14:13 | |
| *** sabeen has joined #openstack-infra | 14:14 | |
| *** jgrimm has joined #openstack-infra | 14:15 | |
| *** markwash has joined #openstack-infra | 14:15 | |
| *** rwsu has joined #openstack-infra | 14:16 | |
| *** hashar has quit IRC | 14:16 | |
| *** rdopieralski has quit IRC | 14:16 | |
| *** unicell has joined #openstack-infra | 14:17 | |
| *** markwash has quit IRC | 14:20 | |
| *** msabramo has quit IRC | 14:21 | |
| *** dkranz has joined #openstack-infra | 14:22 | |
| *** pdmars has quit IRC | 14:23 | |
| *** pdmars has joined #openstack-infra | 14:23 | |
| *** radez is now known as radez_g0n3 | 14:24 | |
| *** radez_g0n3 is now known as radez | 14:27 | |
| *** kgriffs is now known as kgriffs|afk | 14:28 | |
| *** ilyashakhat has quit IRC | 14:28 | |
| *** otherwiseguy has joined #openstack-infra | 14:30 | |
| *** hashar has joined #openstack-infra | 14:33 | |
| *** otherwiseguy has quit IRC | 14:34 | |
| openstackgerrit | Matt Riedemann proposed a change to openstack-infra/elastic-recheck: Add query for glance unit test race bug 1335878 https://review.openstack.org/103556 | 14:36 |
| uvirtbot | Launchpad bug 1335878 in glance "test_download_policy_when_cache_is_not_enabled fails with ECONNREFUSED" [Undecided,Triaged] https://launchpad.net/bugs/1335878 | 14:36 |
| *** msabramo has joined #openstack-infra | 14:36 | |
| *** james_li has joined #openstack-infra | 14:44 | |
| *** craigbr has joined #openstack-infra | 14:47 | |
| *** cnesa has joined #openstack-infra | 14:50 | |
| *** cnesa2 has quit IRC | 14:50 | |
| *** garyh has joined #openstack-infra | 14:51 | |
| *** salv-orlando has quit IRC | 14:54 | |
| *** ildikov has quit IRC | 14:55 | |
| *** zhiyan is now known as zhiyan_ | 14:55 | |
| openstackgerrit | Anita Kuno proposed a change to openstack-infra/config: Creates a list of no-test repositories https://review.openstack.org/102581 | 14:55 |
| *** zhiyan_ is now known as zhiyan | 14:56 | |
| *** NikitaKo_ has joined #openstack-infra | 14:59 | |
| *** zhiyan is now known as zhiyan_ | 14:59 | |
| *** sballe has joined #openstack-infra | 14:59 | |
| *** hdd has quit IRC | 15:00 | |
| openstackgerrit | Michael Krotscheck proposed a change to openstack-infra/config: Modularized StoryBoard Module https://review.openstack.org/98007 | 15:00 |
| *** thedodd has joined #openstack-infra | 15:00 | |
| mordred | fungi: yes - and I was tracking the freespace to make sure we didn't run out | 15:01 |
| krotscheck | Storyboard meeting in #openstack-meeting-3 | 15:02 |
| anteaya | morning mordred, morning krotscheck | 15:02 |
| mordred | fungi: we might want a separate lv - clarkb said we were low on available volume to be able to put there though, so I was trying to re-use some of what we had ... | 15:02 |
| fungi | mordred: awesome. how much room are we going to need for a full pypi mirror, or is that it? if so, maybe we add a 150gb lv for it | 15:03 |
| mordred | fungi: that's it | 15:03 |
| mordred | we now have a full mirror | 15:03 |
| krotscheck | allo allo | 15:03 |
| fungi | yeah, we are getting tight on available cinder quota | 15:03 |
| mordred | fungi: however, a separate lv would allow us to put an FS on it that was less bad at 32k directories | 15:03 |
| fungi | agreed | 15:03 |
| * fungi looks to see what's still open in tha vg | 15:03 | |
| mordred | fungi: before I put the mirror content in that dir, we had 2.3G used | 15:03 |
| mordred | fungi: so we're still double the free space that we'd ever used | 15:04 |
| fungi | yeah, it was basically empty, so we could probably reshuffle it with a little bit of an outage | 15:04 |
| mordred | fungi: and once the full mirror is live, the old partial mirror can go away and we'll be back to 6G free :) | 15:04 |
| fungi | we have 247.95g available in the main vg on static.o.o according to vgs, so we could probably do it without too much impact | 15:04 |
| mordred | oh - neat | 15:05 |
| fungi | wouldn't need to add another cinder volume for it | 15:05 |
| fungi | we could even just grow the /srv/static lv by another ~50g if we wanted with no impact at all, but i'd probably feel a little more comfortable if bandersnatch was mirroring into its own fs anyway | 15:06 |
| *** jgriffit1 is now known as jgriffith | 15:06 | |
| *** craigbr has quit IRC | 15:06 | |
| *** s0nea has joined #openstack-infra | 15:06 | |
| fungi | less likely that it might run away with the volume and impact tarball uploads or something | 15:06 |
| *** marun has joined #openstack-infra | 15:06 | |
| mordred | I'd like another lv - because the 32k dirs thing is crazypants | 15:07 |
| mordred | fungi: any suggestions on which FS would deal with that best | 15:07 |
| mordred | ? | 15:07 |
| mordred | ext4 is ... slow | 15:07 |
| mordred | fungi: for context, /srv/static/pypi/web/simple has 32k direct subdirs | 15:07 |
| fungi | hrm... ext2? | 15:08 |
| *** marun has quit IRC | 15:08 | |
| fungi | it's not like we need much fancy | 15:09 |
| fungi | nor is a journal much help | 15:09 |
| fungi | no requirement for extended attribs... | 15:09 |
| fungi | not going anywhere near the max size limit for ext2 either | 15:09 |
| fungi | but there are certainly other options, i'm just not familiar with some of the more exotic ones | 15:10 |
| *** mrmartin has quit IRC | 15:11 | |
| joa | Small question about the openstack gerrit event stream for CI: Can only the system accounts access it ? | 15:12 |
| *** marun has joined #openstack-infra | 15:12 | |
| *** nati_ueno has joined #openstack-infra | 15:12 | |
| fungi | joa: any account should be able to use the stream-events command | 15:12 |
| *** yfried_ has quit IRC | 15:12 | |
| fungi | we have stream events allowed for the registered users meta-group in gerrit's all-projects acl | 15:13 |
| *** chianingwang has quit IRC | 15:13 | |
| *** marun is now known as marun_afk | 15:14 | |
| joa | okay so I might have an issue with my identity file... thanks | 15:14 |
| *** vhoward- has left #openstack-infra | 15:15 | |
| anteaya | joa: if you are getting an error, please paste the error at paste.openstack.org, and please include the command you ran that resulted in the error | 15:15 |
| anteaya | also paste.openstack.org, is also known as paste.o.o | 15:16 |
| joa | most probably an issue of forgotten password on my side, I get a permission denied (publickey) | 15:16 |
| joa | or bad key uploaded. Just need to scratch the current one and retry | 15:17 |
| *** virmitio has joined #openstack-infra | 15:17 | |
| anteaya | joa: are you trying from your personal account? | 15:17 |
| anteaya | joa: you don't have a ci account yet, correct? | 15:17 |
| *** chandan_kumar has quit IRC | 15:18 | |
| mordred | fungi: actually, ext2/3 has a hard limit of links and you can't make 32k dirs in a single dir | 15:18 |
| *** david-lyle has joined #openstack-infra | 15:18 | |
| joa | yup, trying from personal one | 15:18 |
| mordred | fungi: ext4 bumped the limit | 15:18 |
| *** devkulkarni has joined #openstack-infra | 15:18 | |
| joa | to see how things work; waiting for system account :) | 15:18 |
| *** jcoufal has quit IRC | 15:18 | |
| anteaya | joa: right | 15:19 |
| fungi | mordred: ohhh... good point | 15:19 |
| *** ihrachyshka has quit IRC | 15:19 | |
| anteaya | joa: please paste the command you are trying | 15:19 |
| mordred | fungi: I discovered this the hard way | 15:19 |
| fungi | indeed | 15:19 |
| s0nea | hey infra team... is anyone around who can create an upstream service account for me, please :) | 15:19 |
| fungi | reiserfs is pretty good about that, but... | 15:19 |
| anteaya | joa: you shouldn't have to use a password, unless your ssh key is password protected | 15:20 |
| mordred | wwell, there's the murder | 15:20 |
| anteaya | s0nea: have you read http://ci.openstack.org/third_party.html | 15:20 |
| *** oomichi has joined #openstack-infra | 15:20 | |
| joa | anteaya: it's alright; re-uploading my key fixed that | 15:21 |
| joa | dunnow what kind of mistake I did... | 15:21 |
| *** oomichi has quit IRC | 15:21 | |
| joa | (and yes, my ssh key is passworded) | 15:21 |
| anteaya | joa: hmmm, okay glad you are moving forward again | 15:22 |
| ttx | mordred: infra crew wanted your review o https://review.openstack.org/#/c/102812/ before approving it | 15:22 |
| anteaya | kk | 15:22 |
| anteaya | s0nea: https://review.openstack.org/#/c/102538/3 might be good for you to read as well | 15:22 |
| *** wenlock has joined #openstack-infra | 15:22 | |
| mordred | ttx: looking | 15:22 |
| mordred | fungi: http://pypi.openstack.org/web/simple/ is now available for testing - athough that's not the URL it wants to ultimately be in | 15:22 |
| mordred | fungi: and the first bandersnatch patch should be safe-ish to land-ish | 15:23 |
| *** andreaf_ has quit IRC | 15:23 | |
| mordred | (I mean, it needs to be sheperded in) | 15:23 |
| *** andreykurilin has joined #openstack-infra | 15:24 | |
| anteaya | s0nea: oh and here is a patch on name formating for third party ci accounts: https://review.openstack.org/#/c/101013/ this one is a bit contentious, but that is the patch and comments thus far | 15:24 |
| s0nea | ok, thank anteaya i read http://www.joinfu.com/2014/01/understanding-the-openstack-ci-system/ and http://www.joinfu.com/2014/02/setting-up-an-external-openstack-testing-system/ | 15:24 |
| anteaya | s0nea: great | 15:24 |
| anteaya | s0nea: and I gave you documentation on how to go about requesting a third party ci account | 15:25 |
| anteaya | s0nea: might I ask what it is you plan on testing? | 15:25 |
| *** moted has quit IRC | 15:25 | |
| *** virmitio has left #openstack-infra | 15:25 | |
| *** moted has joined #openstack-infra | 15:25 | |
| *** ildikov has joined #openstack-infra | 15:26 | |
| *** pcm__ has joined #openstack-infra | 15:26 | |
| mordred | ttx: why is there an assumption that there would only be one proposed/ branch? | 15:27 |
| mordred | ttx: (I'm not disagreeing - just looking for more thoughts) | 15:27 |
| pcm__ | hi. Need help with getting latest oslo.messaging release added to global requirements. Can someone advise on the issues I see. | 15:27 |
| ttx | mordred: mostly because it's simpler to pick the "right" one in that case | 15:27 |
| mordred | pcm__: there are issues adding oslo pre-releases to the mirror right now - we're working on a solution ... so you're not alone at least :) | 15:28 |
| ttx | mordred: also the release tooling ensures there is none other when it creates it | 15:28 |
| pcm__ | mordred: I guess that is a good thing :) | 15:28 |
| mordred | ttx: once you release stable/juno, what do you do to proposed/juno? | 15:28 |
| ttx | mordred: it should be deleted | 15:28 |
| ttx | like milestone-proposed was | 15:28 |
| s0nea | anteaya, we developed a cinder driver for our storage system openattic and want to submit it to openstack | 15:29 |
| pcm__ | mordred: Jenkins failed saying that it could not find the release, but it is available. | 15:29 |
| ttx | ideally we'd just "rename" the branch | 15:29 |
| *** otherwiseguy has joined #openstack-infra | 15:29 | |
| fungi | ttx: see comment inline--i think it will break | 15:29 |
| bknudson | pcm__: mordred: is the fix to remove the check for --no-use-mirror in the check-requirements-integration-dsvm ? | 15:29 |
| bknudson | oops, wrong option name | 15:29 |
| mordred | bknudson: nope. it's to completely change our mirroring infrastructure :) | 15:29 |
| bknudson | --no-use-wheel | 15:29 |
| anteaya | s0nea: so Openattic CI? | 15:29 |
| anteaya | s0nea: what is the name of your company? | 15:30 |
| fungi | anteaya: i'm assuming it's this free software project... http://www.open-attic.org/ | 15:31 |
| ttx | fungi: ah! wanted to make it magically compatible with milestone-proposed, but I guess I shouldn't | 15:31 |
| *** timrc-afk is now known as timrc | 15:31 | |
| mriedem | this seems wrong https://review.openstack.org/#/c/83369/2/tox.ini | 15:31 |
| anteaya | s0nea: is fungi correct, is this an opensource project? | 15:31 |
| s0nea | anteaya, yes openattic-ci and http://www.open-attic.org is right | 15:31 |
| *** dangers_away is now known as dangers | 15:32 | |
| anteaya | fungi: should open-attic be in the gate then? | 15:32 |
| *** rcarrill` has quit IRC | 15:32 | |
| fungi | ttx: well, you could then do origin/.*proposed/ | 15:32 |
| *** virmitio has joined #openstack-infra | 15:32 | |
| mordred | ttx, fungi: are we making this merge tags thing too complex? | 15:32 |
| mordred | this runs on a tag event - so we should have the tag name in the ZUUL_ vars | 15:33 |
| *** otherwiseguy has quit IRC | 15:33 | |
| fungi | anteaya: perhaps long-term if it doesn't have any problematic requirements, but short-term doing an external ci system to report on it is also fine (there are other free driver/plugin/backend projects doing similar things with third-party systems already) | 15:33 |
| *** rcarrillocruz has joined #openstack-infra | 15:33 | |
| ttx | mordred: it sounded complicated to me, but hen I was after adding support for proposed/*, not after simplifying it | 15:34 |
| anteaya | fungi: okay thanks, just trying to get the open source stuff into open source | 15:34 |
| mordred | oh - wait | 15:34 |
| fungi | mordred: true, i guess there wouldn't be a tag job called on a tag which exists in multiple branches, so probably fine | 15:34 |
| mordred | it's because we might tag on stable branches | 15:34 |
| mordred | and we don't want to merge those | 15:34 |
| mordred | ok. I remember why we did it this way | 15:34 |
| ttx | mordred: it fails in interestign ways in that specific case | 15:35 |
| fungi | mordred: but we could filter those out as a negation instead if we want | 15:35 |
| *** otherwiseguy has joined #openstack-infra | 15:35 | |
| anteaya | s0nea: great please propose the name open-attic-ci and open-attic-ci for your account name and username (since you have the hyphens in your domain name, let's be consistent) | 15:35 |
| ruhe | anteaya: hi! how can i update contact point entry for murano-ci at https://review.openstack.org/#/admin/groups/91,members ? | 15:35 |
| anteaya | s0nea: once you get up a running, lets look for ways to bring your testing into our openstack infrastructre system | 15:35 |
| *** NikitaKo_ has quit IRC | 15:37 | |
| *** NikitaKo_ has joined #openstack-infra | 15:37 | |
| s0nea | anteaya, don't we need the account for setting up the system? | 15:37 |
| mordred | fungi: so - back to mirror - I think we should go forward with this stuff in place as it is, then make an lv for the stuff and take a little downtime over the weekend to do the volume swap | 15:37 |
| *** basha has quit IRC | 15:38 | |
| anteaya | ruhe: I just created this page for you: https://wiki.openstack.org/wiki/ThirdPartySystems/murano-ci | 15:38 |
| fungi | mordred: how about we add a new lv first, move that new directory tree into it, then the swap becomes a lot faster this weekend? | 15:38 |
| anteaya | ruhe: please fill it out follow this example: https://wiki.openstack.org/wiki/ThirdPartySystems/IBMPowerKVMCI | 15:38 |
| *** msabramo has quit IRC | 15:38 | |
| anteaya | ruhe: once you are finished please email openstack-infra ml with your desired contact changes and the link to your wiki page with your updated information | 15:39 |
| ruhe | anteaya: will do. thank you! | 15:39 |
| anteaya | ruhe: thank you | 15:39 |
| *** Guest9812 has quit IRC | 15:40 | |
| anteaya | s0nea: yes you need to the account to hook into our gerrit | 15:40 |
| anteaya | s0nea: you need to send an email to the openstack-infra ml | 15:40 |
| *** basha has joined #openstack-infra | 15:40 | |
| s0nea | anteaya, ok, now it's clear to me. thank you for your help | 15:41 |
| pcm__ | mordred: On the oslo.messaging pre-release, is there a workaround or a plan to resolve yet? | 15:41 |
| ttx | fungi: so we could just early exit on "git branch --contains $TAG | grep stable" ? | 15:41 |
| anteaya | s0nea: thanks for taking the time to discuss it with me | 15:41 |
| anteaya | s0nea: welcome to openstack | 15:41 |
| anteaya | s0nea: have you spent much time contributing yet? | 15:41 |
| mordred | fungi: well, we need to do a tiny downtime to land the first bandersnatch patch - which is "/etc/init.d/apache stop ; mv /srv/static/pypi/open* /srv/static/pypi/web ; /etc/init.d/apache start" | 15:41 |
| mordred | fungi: if those are on different volumes, that mv will take longer | 15:42 |
| mordred | if they're on the same volume, that shoudl be an almost instant hiccup | 15:42 |
| anteaya | s0nea: the more you understand the developer workflow, the better your default decisions will be regarding your third party ci | 15:42 |
| mordred | fungi: otoh - the openstack dirs don't change nearly as frequently ... so I could just duplicate them on the next volume | 15:43 |
| mordred | fungi: and the above swap could just be the remount | 15:43 |
| fungi | ttx: we could either have it fail early if the tag is contained in either master or stable/.*, or have it only proceed when contained in origin/.*proposed/.* | 15:43 |
| fungi | ttx: i'm mostly indifferent | 15:44 |
| mordred | fungi: actually, I think I like that, an apache restart and umount/mount will be super quick | 15:44 |
| mordred | fungi: actually, we could probably do a umount/mount/graceful and have next to no downtime at all | 15:44 |
| fungi | mordred: i assumed the "move" during the downtime would be mounting the new docroot to a mountpoint which was the same as the original docroot's directory (or change the docroot in the apache vhost slightly)? | 15:45 |
| *** beekneemech is now known as bnemec | 15:45 | |
| mordred | fungi: both | 15:45 |
| *** basha has quit IRC | 15:45 | |
| mordred | the apache vhost docroot needs to shift a directory level so taht we can have the new and old mirrors side by side on the right urls | 15:45 |
| *** hcc is now known as hdd | 15:46 | |
| fungi | yeah, and that gets us both a dedicated logical volume and access to both trees with a very brief (or no) blip, so seems optimal | 15:46 |
| *** nati_ueno has quit IRC | 15:47 | |
| *** cindyo has joined #openstack-infra | 15:47 | |
| mordred | k. I'll work towards that | 15:48 |
| *** Ajaeger has joined #openstack-infra | 15:48 | |
| mordred | dstufft: what FS do you guys use on your mirrors? | 15:48 |
| Ajaeger | Hi infra team, I would appreciate a second +2 for https://review.openstack.org/#/c/98252/ today for some further patches. Thanks! | 15:48 |
| s0nea | s0nea, thanks :) i'm new to such a big open source project and i'm very fascinated of such a organized infrastructure. but it's hard to understand how a project, the developer workflow, "works" | 15:49 |
| dstufft | mordred: dunno, ext4 maybe? | 15:50 |
| dstufft | I'd have to look | 15:50 |
| mordred | dstufft: k. mainly asking because the autoindex seems sluggish | 15:50 |
| anteaya | s0nea: I understand | 15:50 |
| anteaya | s0nea: and yes it is overwhelming | 15:50 |
| anteaya | s0nea: be sure to find me and ask questions as you proceed, I'm happy to help | 15:51 |
| mordred | fungi: anything special other than just lvcreate I should be doing for lvcreating? | 15:51 |
| anteaya | and have done some tutoring for interested parties, you can read the logs from openstack-infra channel at logs.openstack.org | 15:51 |
| *** om has joined #openstack-infra | 15:51 | |
| anteaya | s0nea: look for logs from last tuesday and wednesday from openstack-dev channel | 15:52 |
| *** zzelle_ has joined #openstack-infra | 15:52 | |
| *** gyee has joined #openstack-infra | 15:52 | |
| mordred | fungi: lvcreate main --size 150G --name pypi | 15:52 |
| anteaya | s0nea: search on my name and you should find some logs of me helping some folks learn how to review patches | 15:52 |
| anteaya | s0nea: let me know if I haven't given you enough to find what I mean | 15:52 |
| *** msabramo has joined #openstack-infra | 15:53 | |
| openstackgerrit | Matt Riedemann proposed a change to openstack-infra/elastic-recheck: Add query for glance unit test race bug 1335878 https://review.openstack.org/103556 | 15:53 |
| uvirtbot | Launchpad bug 1335878 in glance "test_download_policy_when_cache_is_not_enabled fails with ECONNREFUSED" [Undecided,Triaged] https://launchpad.net/bugs/1335878 | 15:53 |
| *** james_li has quit IRC | 15:53 | |
| *** gangil has joined #openstack-infra | 15:54 | |
| s0nea | anteaya,thank you, thats nice :) | 15:54 |
| *** denis_makogon has joined #openstack-infra | 15:54 | |
| *** jgallard has quit IRC | 15:55 | |
| anteaya | s0nea: anytime | 15:55 |
| anteaya | Ajaeger: +1 | 15:55 |
| fungi | mordred: yeah, lvcreate and then mkfs... http://ci.openstack.org/sysadmin.html#creating-a-new-logical-volume | 15:55 |
| Ajaeger | anteaya: thanks for the review | 15:56 |
| anteaya | Ajaeger: anytime, I always learn so much from your patches | 15:56 |
| anteaya | :D | 15:56 |
| anteaya | krtaylor: you around yet? | 15:57 |
| anteaya | krtaylor: feel like taking an initial pass at the meeting agenda? | 15:57 |
| anteaya | krtaylor: I'm going to be on the road the next three mondays after today | 15:57 |
| dstufft | mordred: fwiw I plan on someday changing the FS structure so it uses hashes in the URLs like a/b/c/d/abcdefgasdadsad/Dajngo-1.0.tar.gz | 15:58 |
| *** james_li has joined #openstack-infra | 15:58 | |
| anteaya | krtaylor: I'd like to start getting you into the chair chair for third party meetings, now that we have a bit of structure to work with | 15:58 |
| anteaya | krtaylor: if you are okay with that | 15:58 |
| Ajaeger | anteaya: you learn from mine? That surprise me. Thanks! | 15:58 |
| anteaya | Ajaeger: oh yes, always | 15:59 |
| Ajaeger | fungi: thanks for the review! | 15:59 |
| cindyo | anteaya: using weechat per your suggestion! Thanks! | 15:59 |
| *** chandan_kumar has joined #openstack-infra | 15:59 | |
| anteaya | Ajaeger: your patches are well documented in the commit message, and I like your motivation for your patches | 15:59 |
| anteaya | cindyo: hey there | 15:59 |
| anteaya | cindyo: yay! | 15:59 |
| anteaya | cindyo: much less noise from your client than last week | 15:59 |
| *** praneshp has joined #openstack-infra | 16:00 | |
| *** blamar has joined #openstack-infra | 16:00 | |
| anteaya | and weechat is a popular choice in channel, so if you need a hand finding a command, do ask | 16:00 |
| cindyo | anteaya: yes! Still not sure about multiple channels but at least this won't disconnect a lot | 16:00 |
| Ajaeger | anteaya: seems I don't review as many infra patches as you do ;). Will try to keep the level up ;) | 16:00 |
| anteaya | cindyo: sure, that makes sense, a bit at a time | 16:00 |
| *** thedodd has quit IRC | 16:00 | |
| anteaya | cindyo: just type /join #openstack-dev in the text line | 16:00 |
| *** marun_afk is now known as marun | 16:00 | |
| anteaya | and hit enter | 16:00 |
| *** NikitaKo_ has quit IRC | 16:00 | |
| anteaya | you should join the -dev channel as well | 16:01 |
| *** pblaho has quit IRC | 16:02 | |
| anteaya | cindyo: did that work? | 16:02 |
| cindyo | anteaya: cool!! | 16:02 |
| anteaya | cindyo: okay so escape key and digit work for channels 1 through 10 | 16:02 |
| anteaya | join as many as you like | 16:02 |
| fungi | or alt+digit | 16:03 |
| anteaya | escape key, letter j and two digits work for channels 11 to 99 | 16:03 |
| anteaya | yay, thanks fungi | 16:03 |
| fungi | or alt+j,digit,digit | 16:03 |
| cindyo | anteaya: would these two topics be the best to hang out on , since we are stackforge? | 16:03 |
| anteaya | not sure how to get to channels over 2 digits | 16:03 |
| cindyo | anteaya: channels I mean | 16:03 |
| fungi | anteaya: /buffer 1234 | 16:03 |
| *** andreykurilin has quit IRC | 16:03 | |
| *** james_li_ has joined #openstack-infra | 16:04 | |
| zaro | morning | 16:04 |
| anteaya | cindyo: sure, and #openstack-meeting, #openstack-meeting-alt and #openstack-meeting-3 | 16:04 |
| anteaya | morning zaro | 16:04 |
| anteaya | fungi: ah cool, thank you | 16:04 |
| anteaya | cindyo: see, by teaching you, I get to learn something new | 16:04 |
| anteaya | :D | 16:04 |
| anteaya | yaya | 16:04 |
| cindyo | anteaya: yeah sure, I think you know it all | 16:04 |
| *** anode has joined #openstack-infra | 16:05 | |
| *** reed_ has joined #openstack-infra | 16:05 | |
| cindyo | anteaya: :) | 16:05 |
| anteaya | cindyo: you will soon find out I don't | 16:05 |
| anteaya | I just ask a lot of questions | 16:05 |
| cindyo | anteaya: haha | 16:05 |
| anteaya | cindyo: :D | 16:05 |
| cindyo | anteaya: that's what I do too | 16:05 |
| anteaya | good | 16:05 |
| anteaya | you are in a good place then | 16:06 |
| anteaya | look at your weechat | 16:06 |
| anteaya | I just pinged you in -dev | 16:06 |
| *** UtahDave has joined #openstack-infra | 16:06 | |
| *** james_li has quit IRC | 16:06 | |
| *** james_li_ is now known as james_li | 16:06 | |
| anteaya | the 3 should be highlighted | 16:06 |
| anteaya | go to window 3 and reply | 16:06 |
| fungi | ipse se nihil scire id unum sciat | 16:06 |
| cindyo | hmmm 3 is not highlighted | 16:06 |
| anteaya | hmmmm | 16:07 |
| anteaya | maybe you don't have names highlighted then | 16:07 |
| anteaya | ah maybe your name isn't registered? | 16:07 |
| *** Nikolay_St has joined #openstack-infra | 16:07 | |
| Nikolay_St | hi all | 16:07 |
| anteaya | cindyo: do you know about registering your name with freenode | 16:07 |
| *** praneshp has quit IRC | 16:07 | |
| anteaya | cindyo: ping | 16:07 |
| Nikolay_St | we have this error in blazar: http://logs.openstack.org/39/102339/4/check/gate-blazar-python26/5d74e3c/console.html#_2014-06-30_15_26_03_956 | 16:07 |
| Nikolay_St | any suggestions? | 16:08 |
| openstackgerrit | Andreas Jaeger proposed a change to openstack-infra/config: Adds a docs-specs repository https://review.openstack.org/103115 | 16:08 |
| anteaya | cindyo: try this set of instructions: http://www.wikihow.com/Register-a-User-Name-on-Freenode | 16:08 |
| anteaya | cindyo: just follow the command line stuff | 16:08 |
| cindyo | anteaya: at the bottom of weechat I saw a pink number, does that mean go to the channel? | 16:09 |
| mordred | fungi: actually - screw it - I'm just going to make the downtime be a single apache graceful | 16:09 |
| fungi | mordred: wfm! | 16:09 |
| openstackgerrit | A change was merged to openstack-infra/elastic-recheck: Add query for glance unit test race bug 1335878 https://review.openstack.org/103556 | 16:09 |
| uvirtbot | Launchpad bug 1335878 in glance "test_download_policy_when_cache_is_not_enabled fails with ECONNREFUSED" [Undecided,Triaged] https://launchpad.net/bugs/1335878 | 16:09 |
| *** thedodd has joined #openstack-infra | 16:09 | |
| fungi | Nikolay_St: bug 1282795 | 16:09 |
| uvirtbot | Launchpad bug 1282795 in openstack-ci "InterpreterNotFound: python2.6" [Low,Confirmed] https://launchpad.net/bugs/1282795 | 16:09 |
| anteaya | cindyo: yes | 16:10 |
| anteaya | cindyo: yes the pink number says, soneone in that channel is alerting you | 16:10 |
| Nikolay_St | fungi: uvirtbot: thx, guys | 16:10 |
| anteaya | cindyo: so you can have over a hundred channel connections, and if someone calls you, you can see | 16:10 |
| cindyo | anteaya: ahh! I'll read the link, thanks! | 16:10 |
| *** rcarrillocruz has quit IRC | 16:11 | |
| anteaya | cindyo: awesome | 16:11 |
| anteaya | cindyo: yes best to register your nickname with freenode | 16:11 |
| anteaya | I have to get a walk in before the third party meeting | 16:11 |
| anteaya | back in a bit | 16:11 |
| fungi | anteaya: cindyo: i have some more important channels set to track activity (so i know when anyone in that channel says something) and others set to only highlight when someone says a particular keyword (for example, my nick/name or other topics of interest) | 16:11 |
| openstackgerrit | A change was merged to openstack-infra/config: Create zuul template translation-jobs https://review.openstack.org/98252 | 16:12 |
| fungi | keeps the volume of channels manageable | 16:12 |
| *** jpich has quit IRC | 16:12 | |
| *** praneshp has joined #openstack-infra | 16:13 | |
| cindyo | anteaya: cool! walking and exercise is the best for clearing our minds! | 16:13 |
| *** IvanBerezovskiy has left #openstack-infra | 16:13 | |
| mordred | fungi: any issues with me using dir_index on this FS? | 16:13 |
| *** doddstack has joined #openstack-infra | 16:15 | |
| *** thedodd has quit IRC | 16:15 | |
| jeblair | mordred: what fs? | 16:16 |
| *** rcarrillocruz has joined #openstack-infra | 16:16 | |
| fungi | mordred: reading up on it | 16:17 |
| fungi | jeblair: ext4 | 16:17 |
| jeblair | mordred: because if it's something we care about, i'm surprised it's not already ext4 | 16:17 |
| jeblair | which has that built in | 16:17 |
| fungi | ahh, yep. on by default in 4, not in 3 | 16:18 |
| jeblair | fungi, mordred: i didn't mean which filesystem type, i mean, literally, which actual filesystem is this discussion about? | 16:18 |
| fungi | jeblair: new filesystem for pypi.o.o docroot | 16:18 |
| jeblair | oh, yeah, so it should be created as ext4 | 16:18 |
| mordred | jeblair: for the bandersnatch mirror | 16:18 |
| jeblair | and setting dir_index should not be necessary | 16:18 |
| mordred | and yeah. it will be ext4 | 16:18 |
| mordred | no? | 16:18 |
| fungi | jeblair: since bandersnatch will mirror ~100g of data, this helps keep it from possibly starving out the fs where tarballs are being uploaded and such | 16:19 |
| *** arnaud_ has joined #openstack-infra | 16:19 | |
| mordred | jeblair: did you see the piece about the full mirror having a dir with >32K subdirs? this is the thing I was hoping to optimize for with dir_index - but if you dont' think it's necessary, I'm happy to not do | 16:19 |
| fungi | er, well, i guess not tarballs specifically, since we already moved that onto its own fs as well | 16:19 |
| mordred | oh - I see - built in | 16:19 |
| mordred | sorry - didn't read enough words above | 16:19 |
| dstufft | mordred: looks like we're using ext4 for our mirro btw | 16:20 |
| dstufft | I just looked | 16:20 |
| *** pafuent has joined #openstack-infra | 16:20 | |
| fungi | but could potentially break status, releasestatus, reviewday, et cetera if it runs wild | 16:20 |
| dstufft | Also we're currently using 93GB of data | 16:20 |
| openstackgerrit | Andreas Jaeger proposed a change to openstack-infra/config: Add gating for security-doc repo https://review.openstack.org/103578 | 16:20 |
| fungi | er, not status either... i guess we moved that to another host | 16:20 |
| fungi | this /srv/static lv is becoming increasingly unused ;) | 16:20 |
| fungi | perhaps we should move /srv/static itself back to / | 16:21 |
| *** jedimike has quit IRC | 16:21 | |
| fungi | but that *will* be downtime for logs and such, to move their mountpoints to the root fs | 16:22 |
| fungi | so maybe something for later | 16:22 |
| *** hashar has quit IRC | 16:22 | |
| mordred | yah. doesn't seem super important | 16:22 |
| mordred | dstufft: thanks! | 16:22 |
| mordred | dstufft: and yes - I got that same figure :) | 16:22 |
| dstufft | give yourself lots of room | 16:22 |
| dstufft | that's apparently growing quicker than I thought | 16:22 |
| dstufft | couple G a month or so it appears | 16:23 |
| fungi | well, projects have a tendency to release things, and not un-release things, so makes sense ;) | 16:23 |
| *** doddstack has quit IRC | 16:23 | |
| *** thedodd has joined #openstack-infra | 16:24 | |
| openstackgerrit | A change was merged to openstack-dev/pbr: Allow examining parsing exceptions. https://review.openstack.org/80856 | 16:27 |
| *** marcoemorais has joined #openstack-infra | 16:30 | |
| *** andreaf_ has joined #openstack-infra | 16:30 | |
| *** morganfainberg_Z is now known as morganfainberg | 16:30 | |
| *** arnaud_ has quit IRC | 16:32 | |
| *** markwash has joined #openstack-infra | 16:34 | |
| Ajaeger | anteaya: I hope you like https://review.openstack.org/103578 as well ;) | 16:34 |
| *** james_li has quit IRC | 16:37 | |
| cindyo | fungi: thanks for your tips too, I had to page up to get it all. Not as fast as you guys in the IRC yet! | 16:39 |
| openstackgerrit | Andreas Jaeger proposed a change to openstack-infra/config: Add gating for security-doc repo https://review.openstack.org/103578 | 16:39 |
| openstackgerrit | Monty Taylor proposed a change to openstack-infra/config: Stop running pypi-mirror for old mirror creation https://review.openstack.org/103360 | 16:41 |
| openstackgerrit | Monty Taylor proposed a change to openstack-infra/config: Start building a bandersnatch mirror https://review.openstack.org/103353 | 16:41 |
| openstackgerrit | Monty Taylor proposed a change to openstack-infra/config: Make select-mirror use the bandersnatch mirror https://review.openstack.org/103354 | 16:41 |
| openstackgerrit | Monty Taylor proposed a change to openstack-infra/config: Delete select-mirror and all the args https://review.openstack.org/103355 | 16:41 |
| mordred | fungi: ^^ | 16:41 |
| *** andreaf_ has quit IRC | 16:41 | |
| mordred | fungi: so the plan in my head is - I'm currently moving the bandersnatch dir into the /srv/static/mirror volume | 16:41 |
| openstackgerrit | Andreas Jaeger proposed a change to openstack-infra/config: Fix django_openstack_auth translation setup https://review.openstack.org/96132 | 16:41 |
| mordred | fungi: then I'll _copy_ the openstack dirs into place | 16:42 |
| *** yfried_ has joined #openstack-infra | 16:42 | |
| mordred | fungi: so all we need to do is land the first patch in the sequence and graceful apache and it _should_ just work with no interruption in service | 16:42 |
| mordred | once the new mirror is happy and also we're still serving the old mirror from the new location | 16:42 |
| *** dkehn__ has joined #openstack-infra | 16:42 | |
| mordred | we can delete /srv/static/pypi at our leisure | 16:42 |
| mordred | jeblair: ^^ | 16:43 |
| mordred | btw - I have verified that followup bandernsatch runs take very little time - so the 5 minute cron should work very well | 16:43 |
| mordred | and could probably go to a 1 or 2 minute cron easily | 16:44 |
| *** sarob has joined #openstack-infra | 16:44 | |
| jeblair | neatos | 16:45 |
| *** dkehnx has quit IRC | 16:45 | |
| mordred | jeblair: if you haven't looked at the last two changes in that sequence yet - they're fun - they involve a lot of deleting of things | 16:45 |
| *** dstanek is now known as dstanek_zzz | 16:46 | |
| *** miqui has quit IRC | 16:46 | |
| * mordred expects the mv to take about another hour | 16:47 | |
| *** melwitt has joined #openstack-infra | 16:47 | |
| mordred | flaper87: moving here | 16:47 |
| * flaper87 knocks infra's doors | 16:47 | |
| mordred | jeblair, fungi: so - ... flaper87 is having a problem with tagging a glanceclient release | 16:48 |
| flaper87 | this is the last build: https://jenkins02.openstack.org/job/python-glanceclient-tarball/3/console | 16:48 |
| flaper87 | It creates a tarball using the wrong version string and I'm not sure why | 16:48 |
| *** NikitaKo_ has joined #openstack-infra | 16:48 | |
| flaper87 | The tag I'm trying to release is 0.13.0 and it's a signed tag | 16:48 |
| mordred | there are two issues ... | 16:49 |
| flaper87 | I tried running sdist locally and it worked perfectly. AFAIK mordred tried to do the same and it worked for him as well | 16:49 |
| mordred | 1) it allowed him to push an unsigned tag this morning | 16:49 |
| flaper87 | oh yeah, I was able to push an unsigned tag | 16:49 |
| flaper87 | bad gerrit, bad gerrit. | 16:49 |
| mordred | which is, I _think_ maybe because there is still a create acl in the acls file? | 16:49 |
| *** devkulkarni1 has joined #openstack-infra | 16:49 | |
| mordred | so maybe we should make a pass and clean those up | 16:49 |
| *** devkulkarni has quit IRC | 16:49 | |
| jeblair | fungi: is that in your cleanup series? | 16:50 |
| *** sarob has quit IRC | 16:51 | |
| fungi | jeblair: yep, https://review.openstack.org/93067 | 16:51 |
| * mordred goes to look ... | 16:51 | |
| *** sarob has joined #openstack-infra | 16:51 | |
| *** NikitaKo_ has quit IRC | 16:53 | |
| krtaylor | anteaya, yes and yes! | 16:53 |
| *** craigbr has joined #openstack-infra | 16:53 | |
| *** basha has joined #openstack-infra | 16:53 | |
| krtaylor | anteaya, I'll be happy to chair next week | 16:53 |
| fungi | that series insta-conflicts when any new config changes are approved, so if we can get some consensus that we're okay with them, i'll update them one last time immediately prior to sending into the gate | 16:53 |
| jeblair | fungi, mordred: maybe we should do that right now? | 16:54 |
| fungi | (well, not when any new config changes are approved, but fairly rapidly nevertheless) | 16:54 |
| *** CaptTofu has quit IRC | 16:54 | |
| mordred | fungi: ++ | 16:54 |
| mordred | I'm reading through them now - but I think insta-approving a new set would be great | 16:55 |
| mordred | (they look great so far) | 16:55 |
| *** james_li has joined #openstack-infra | 16:55 | |
| *** annegent_ has joined #openstack-infra | 16:55 | |
| *** anode has quit IRC | 16:56 | |
| james_li | Hi All, we are seeing fatal: unable to connect to git.openstack.org http://logs.openstack.org/34/102634/3/check/gate-solum-python27/c297c77/console.html | 16:57 |
| *** SpamapS_ is now known as SpamapS | 16:57 | |
| flaper87 | fungi: jeblair mordred do you think the problem with the tarball job is related to this cleanup? Any other ideas about what it could be? | 16:57 |
| *** andreykurilin has joined #openstack-infra | 16:57 | |
| *** SpamapS has quit IRC | 16:57 | |
| *** SpamapS has joined #openstack-infra | 16:57 | |
| fungi | flaper87: that was the run from the unsigned tag? | 16:57 |
| mordred | fungi: I think zaro was saying "why isn't requireChangeId in global ... not the CLA one | 16:58 |
| *** sarob has quit IRC | 16:58 | |
| flaper87 | fungi: nope, that one is the signed tag | 16:58 |
| mordred | fungi: https://review.openstack.org/#/c/93064/3/doc/source/jeepyb.rst | 16:58 |
| *** anode has joined #openstack-infra | 16:58 | |
| fungi | flaper87: did you previously push an unsigned tag with that same number? | 16:58 |
| mordred | fungi: and I agree - requireChangeId should be global | 16:58 |
| flaper87 | fungi: yes | 16:58 |
| *** sarob has joined #openstack-infra | 16:58 | |
| flaper87 | fungi: The previous tag was deleted and then I pushed the signed one | 16:58 |
| fungi | mordred: oh, well yes requireChangeId probably should be global. that's one we can add a cleanup pass for if we add it to all-projects first | 16:59 |
| mordred | fungi: yah. I don't think we need to chagne that now obviously | 16:59 |
| fungi | flaper87: that's the problem. tag deletion isn't particularly effective because updates won't spot that a tag was replaced | 17:00 |
| fungi | flaper87: you'll need to tag and relase a 0.14.0 or 0.13.1 or 0.13.0.1 or something newer than the bad tag | 17:00 |
| flaper87 | fungi: ooook, lemme do that | 17:01 |
| *** devkulkarni1 has quit IRC | 17:01 | |
| mordred | fungi: btw - I'm super happy about the tab normalization | 17:01 |
| fungi | mordred: yeah, that was a happy discovery | 17:01 |
| mordred | the git leading tab in ini file thing is about the most ass-tastic thing ever | 17:01 |
| *** yfried_ has quit IRC | 17:01 | |
| fungi | mordred: i expect it to significantly reduce errors in new/changed acls | 17:02 |
| fungi | mordred: however, we *do* need to keep in mind that these still can't be parsed correctly by python's configparser module because option keys can be duplicated within a section | 17:02 |
| *** craigbr has quit IRC | 17:03 | |
| *** NikitaKo_ has joined #openstack-infra | 17:03 | |
| fungi | (food for thought if we decide to make a linter of some sort for them down the road) | 17:03 |
| mordred | fungi: nod | 17:03 |
| mordred | fungi: it would be neat if python's config parser had an option called "read git config files" | 17:03 |
| *** yfried_ has joined #openstack-infra | 17:04 | |
| fungi | indeed | 17:04 |
| mordred | also, if there were an option called "allow mysql style no-equal-sign toggle flags" | 17:04 |
| mordred | which lets you do things like [mysqld]\nuse-innodb | 17:04 |
| fungi | that^2 | 17:04 |
| mordred | and have it show up as config['mysql']['use-innodb']==True | 17:04 |
| *** derekh_ has quit IRC | 17:05 | |
| mordred | dhellmann: ^^ you fix python all the time, right? | 17:05 |
| *** lttrl has quit IRC | 17:06 | |
| *** NikitaK__ has joined #openstack-infra | 17:07 | |
| *** NikitaKo_ has quit IRC | 17:07 | |
| *** NikitaK__ has quit IRC | 17:11 | |
| mordred | fungi: do you want me to finish reviewing the cleanup patches before you respin? | 17:11 |
| clarkb | o/ | 17:11 |
| *** jlibosva has quit IRC | 17:12 | |
| fungi | mordred: or we can approve the first few in the series if we see gains from them and land the later ones another time | 17:12 |
| mordred | fungi: I think we should land them all | 17:12 |
| anteaya | morning clarkb | 17:12 |
| mordred | fungi: and I am +2 on all of them | 17:12 |
| fungi | mordred: okay, cool. i'll update them now | 17:13 |
| anteaya | I haven't had a chance to review all fungi's changes | 17:13 |
| anteaya | but I am +1 on the design decision behind them all | 17:13 |
| mordred | anteaya: you'll like them - they simplify | 17:13 |
| mordred | anteaya: and are quite pleasing to the eye | 17:13 |
| anteaya | if they go through before I can +1 them individually | 17:13 |
| mordred | clarkb: morning! | 17:13 |
| *** NikitaKo_ has joined #openstack-infra | 17:13 | |
| fungi | i just like them because they delete lines. i like deleting things | 17:13 |
| fungi | a lot | 17:13 |
| mordred | clarkb: morning update - I'm close to having the bandersnatch mirror ready to go and will want to land the first patch of the series in about an hour | 17:14 |
| clarkb | good morning | 17:14 |
| anteaya | mordred: yes, I like them, haven't been able to review every file yet though | 17:14 |
| anteaya | and don't want to hold things up if today is the day they merge | 17:14 |
| clarkb | mordred: cool, so quick question. Are you replacing the existing mirror at the same time? | 17:14 |
| anteaya | merge them by all means | 17:14 |
| mordred | clarkb: so if you don't mind eyeballing at least the first one https://review.openstack.org/103353 I'd appreciate it | 17:14 |
| mordred | clarkb: side by side | 17:14 |
| clarkb | mordred: because I am curious what you think about the wheels and enforcement piece of this | 17:14 |
| *** nati_ueno has joined #openstack-infra | 17:14 | |
| mordred | clarkb: what do you mean? | 17:15 |
| mordred | clarkb: for enforcement, I think we're doing a great job with the requirements repo, so I don't think we need enforcement | 17:15 |
| *** pabelanger_ is now known as pabelanger | 17:15 | |
| *** pabelanger has joined #openstack-infra | 17:15 | |
| *** annegent_ has quit IRC | 17:15 | |
| *** gokrokve has joined #openstack-infra | 17:15 | |
| mordred | clarkb: for wheels - I think we _do_ want to build wheels of things for our target hosts - but probably want to rework that since it wasn't being used in the gate yet anyway, so I haven't done any hacking on it yet | 17:16 |
| clarkb | mordred: ok I wasn't sure if there are missing pieces of encforcement. | 17:16 |
| mordred | clarkb: I don't think so - I think we're pretty solid on it | 17:16 |
| clarkb | mordred: mostly thinking on the unittest side where we don't really enforce much | 17:16 |
| mordred | clarkb: we don't - but we enforce on the repos ... the main thing that could slip by there is someone putting reqs directly into tox.ini | 17:16 |
| *** harlowja has joined #openstack-infra | 17:17 | |
| *** andreykurilin has quit IRC | 17:17 | |
| mordred | but I think if we see that become a problem, we can just make a job that reads tox.ini and interprets the reqs value - I think socially cores are unlikely to approve such a thing at this point | 17:17 |
| clarkb | mordred: fungi: for the thing I want to tackle today (ssd volume on elasticsearch07) any opposition to me just deleting ES01 and its volume rather than going through and cleaning them up in a nice way? | 17:17 |
| jeblair | clarkb: es01 or es07? | 17:18 |
| *** mmaglana has joined #openstack-infra | 17:18 | |
| fungi | clarkb: umm... what would you have had in mind for "a nice way"? | 17:18 |
| clarkb | jeblair: es01 is still around and is consuming 1TB ofvolume. so I would delete 01 | 17:18 |
| clarkb | fungi: well I can unmount the fs, remove the physical device, delete the volume then delete the host | 17:18 |
| fungi | oh | 17:18 |
| clarkb | and stop at some point in the middle if necessary for other raisin | 17:18 |
| fungi | no, no point at all in being nice to a server you're going to delete anyway | 17:19 |
| mordred | ++ | 17:19 |
| jeblair | clarkb: why are you deleting es01? | 17:19 |
| clarkb | jeblair: because it isn't doing anything but consuming quota | 17:19 |
| fungi | es01 has been out of service for a week or more (replaced by 07) | 17:19 |
| jeblair | is 07 new? | 17:19 |
| clarkb | yes | 17:20 |
| *** andreykurilin has joined #openstack-infra | 17:20 | |
| jeblair | oh, you're rolling hosts; i thought you were rolling volumes on hosts | 17:20 |
| mordred | jeblair: 7 is the new black | 17:20 |
| clarkb | jeblair: well we tried rolling hosts first | 17:20 |
| *** Guest9812 has joined #openstack-infra | 17:20 | |
| clarkb | jeblair: because dstufft had indicated it helped their gluster at one point | 17:20 |
| fungi | that was a host roll from an earlier attempt to work around the slow-volume-problem | 17:20 |
| clarkb | jeblair: however in our case it doesn't seem to have done much | 17:20 |
| dstufft | :[ | 17:20 |
| clarkb | so now the thought is to roll onto ssd volumes | 17:20 |
| clarkb | and see if that helps any | 17:21 |
| mordred | dstufft: there's so much bad in teh scrollback, which thing makes you unhappy this time? :) | 17:21 |
| clarkb | or we can continue to hope pvo shows up (mordred did you ping pvo?) and we have have a deeper discussion about it | 17:21 |
| dstufft | mordred: oh, just that the thing that helped gluster didn't help y'all | 17:21 |
| jeblair | clarkb: is there an end state that has our first es node being called 01? or are we going to be 02-07 for a while? | 17:21 |
| fungi | es01's cinder volume was waaay slower than the other servers, so bringing a new cluster member in and stopping that one was an earlier attempt at solving this... until the problem cropped up on yet another cluster member | 17:21 |
| mordred | oh - crappit | 17:21 |
| clarkb | or submit a ticket but in my experience that tends to not help as much when you need a real conversation around how to make it work | 17:21 |
| mordred | clarkb: can you tell me in small words because of my dumb manager brain what I need to ask? | 17:21 |
| clarkb | jeblair: just 02-07 | 17:21 |
| mordred | clarkb: I've got troy in a channel | 17:21 |
| clarkb | mordred: I think mostly looking for information on rax side how they would build such a thing | 17:22 |
| clarkb | mordred: we are kind of flying blind here when all of a sudden it takes 5 seconds for all disk ops | 17:22 |
| clarkb | mordred: is that a network bandwidth issue? is that a spinning disk issue? | 17:22 |
| *** Sukhdev has joined #openstack-infra | 17:23 | |
| clarkb | is that a particular hypervisor being sad? | 17:23 |
| clarkb | and given whatever it is how should we build a ~6TB elasticsearch cluster that needs reasonable IO and cpu performance | 17:23 |
| openstackgerrit | Andreas Jaeger proposed a change to openstack-infra/config: IRC setup for security-doc and training-guides https://review.openstack.org/103597 | 17:23 |
| *** NikitaKo_ has quit IRC | 17:25 | |
| *** james_li has quit IRC | 17:25 | |
| clarkb | fungi: jeblair: at this point I could also add an ssd volume to 01, and readd it to the cluster (I think this may require firewall updates too) | 17:26 |
| clarkb | then work to kill 07. But again its a bit like flying blind and a bit more of a conversation with the other side would be good | 17:26 |
| openstackgerrit | A change was merged to openstack-infra/config: Update pipelines to new Zuul syntax https://review.openstack.org/102033 | 17:26 |
| *** james_li has joined #openstack-infra | 17:27 | |
| anteaya | krtaylor: any progress on the meeting agenda? | 17:27 |
| fungi | clarkb: i am so far succeeding in not letting the ocd in me care that we have servers numbered starting from 2 instead of 1 (after all, why not start at 0!) since there is no efficiency gain from shuffling 01 back into service and shifting shards back off 07 | 17:28 |
| yolanda | jeblair, mordred, can we tak about nodepool and dib process for building images? | 17:28 |
| clarkb | fungi: ya it doesn't bother me either. Just realized that is an option if others prefer it | 17:28 |
| mordred | jeblair: you were saying in your review that you think we should do the "build once" approach out of the gate | 17:28 |
| mordred | jeblair: but that will necessitate a data model change I think? | 17:29 |
| *** miqui has joined #openstack-infra | 17:29 | |
| mordred | clarkb: ok. I've poked troy | 17:29 |
| *** otherwiseguy has quit IRC | 17:29 | |
| jeblair | mordred, yolanda: i think i left a comment suggesting how it could be done... let me see | 17:29 |
| *** otherwiseguy has joined #openstack-infra | 17:30 | |
| anteaya | krtaylor: sorry I haven't gotten to it yet, I will expand today's meeting agenda | 17:31 |
| openstackgerrit | Jeremy Stanley proposed a change to openstack-infra/config: Normalize Gerrit ACL documentation https://review.openstack.org/93064 | 17:31 |
| openstackgerrit | Jeremy Stanley proposed a change to openstack-infra/config: Add a script to normalize Gerrit ACLs https://review.openstack.org/93065 | 17:31 |
| openstackgerrit | Jeremy Stanley proposed a change to openstack-infra/config: Strip/condense whitespace and sort Gerrit ACLs https://review.openstack.org/93066 | 17:31 |
| openstackgerrit | Jeremy Stanley proposed a change to openstack-infra/config: Get rid of unneeded tags create in Gerrit ACLs https://review.openstack.org/93067 | 17:31 |
| openstackgerrit | Jeremy Stanley proposed a change to openstack-infra/config: Remove active project.state in Gerrit ACLs https://review.openstack.org/93068 | 17:31 |
| openstackgerrit | Jeremy Stanley proposed a change to openstack-infra/config: Strip *.owner Administrators from Gerrit ACLs https://review.openstack.org/93069 | 17:31 |
| openstackgerrit | Jeremy Stanley proposed a change to openstack-infra/config: Sort exclusiveGroupPermissions in Gerrit ACLs https://review.openstack.org/93070 | 17:31 |
| openstackgerrit | Jeremy Stanley proposed a change to openstack-infra/config: Combine infra/ci core/admin gerrit groups https://review.openstack.org/102271 | 17:31 |
| *** james_li has quit IRC | 17:32 | |
| *** UtahDave has left #openstack-infra | 17:32 | |
| cindyo | anteaya: do you usually '/part' or '/close' the buffer when you want to leave weechat? | 17:32 |
| *** rcarrill` has joined #openstack-infra | 17:32 | |
| *** james_li has joined #openstack-infra | 17:32 | |
| jeblair | mordred: we're much closer to the data model necessary since you first started working on it (we're also no longer in a time crunch since we have trusty) | 17:33 |
| krtaylor | anteaya, I reviewed, did you want me to add anything? | 17:33 |
| jeblair | mordred, yolanda: nodepool now has a 'label' construct, which has an image name | 17:33 |
| openstackgerrit | A change was merged to openstack-dev/pbr: Check for git before querying it for a version https://review.openstack.org/98273 | 17:34 |
| mordred | jeblair: nod | 17:34 |
| clarkb | cindyo: I think most of us run weechat perpetually | 17:34 |
| jeblair | mordred, yolanda: that currently is linked to the provider, but it could reference a new "images" section in the nodepool config | 17:34 |
| *** rcarrillocruz has quit IRC | 17:34 | |
| jeblair | and drop the link to the provider base-image bit | 17:34 |
| clarkb | cindyo: for example my weechat has been running since april 8th (heartbleed reboots) | 17:35 |
| anteaya | cindyo: I have weechat running in tmux, and I detach from tmux | 17:35 |
| mordred | jeblair: yah. that sounds completely reasonable - yolanda - does that make sense to you? | 17:35 |
| anteaya | cindyo: I use /close when I want to close one window | 17:35 |
| anteaya | cindyo: and quit when I want to quit weechat | 17:35 |
| anteaya | cindyo: since I have it running in tmux, I leave it running and just attach and detach | 17:35 |
| *** arnaud_ has joined #openstack-infra | 17:35 | |
| anteaya | krtaylor: yeah, I needed to fix the dates | 17:35 |
| anteaya | krtaylor: don't worry, I'm addressing it | 17:36 |
| anteaya | krtaylor: then you will see what I mean | 17:36 |
| cindyo | thanks. How do you detach from tmux while in the weechat app? | 17:36 |
| jeblair | mordred: i'm sensitive to the idea that this may be a smaller step to get us there... but here's why i was thinking it might be worth a little more investment (assuming we have time): | 17:36 |
| yolanda | mordred, jeblair, so when nodepool is runninng, it looks into the labels/images section, checks if we have an image built for that and if not, it starts to build one | 17:36 |
| cindyo | anteaya: clarkb ^^ | 17:36 |
| anteaya | cindyo: I usually move to the first window, so if I am pinged in channels while I am away I can tell once I return | 17:37 |
| anteaya | then control+b and d for detach | 17:37 |
| jeblair | mordred: when we actually start to use this, and if there's a problem, i'd rather not have to figure out why there is a difference between one of the 8 'precise' images we built | 17:37 |
| anteaya | the control+b is tmux the next command is for you | 17:37 |
| openstackgerrit | pritesh proposed a change to openstack-dev/hacking: Add hacking check for vim configuration in source files https://review.openstack.org/101969 | 17:37 |
| yolanda | mordred, jeblair, and where will we store it? we have an snapshot_image table, but it's linked to the provider | 17:37 |
| jeblair | mordred: and when we make changes, i'd rather not have to worry about the nodepool host building 8x the number of images (and how long/how much resources that will take) | 17:38 |
| yolanda | shall we just use the same, and leave provider blank? | 17:38 |
| *** adrian_otto has joined #openstack-infra | 17:38 | |
| jeblair | yolanda: i think we still need a snapshot images table because there will be a glance id for each image uploaded to a provider | 17:38 |
| jeblair | yolanda: so that stays the same | 17:39 |
| openstackgerrit | Davanum Srinivas (dims) proposed a change to openstack-infra/config: Create git repo and jenkins jobs for oslo.utils https://review.openstack.org/101230 | 17:39 |
| mordred | jeblair: makes sense - totally agree | 17:39 |
| cindyo | hmmm control b + d not working from my mac keyboard | 17:39 |
| *** cindyo has quit IRC | 17:39 | |
| yolanda | jeblair, so we create the image once, then we upload to glance for all the providers needed? | 17:39 |
| pleia2 | fire alarm testing in my building today, my favorite ;) | 17:39 |
| *** tkelsey_ has quit IRC | 17:40 | |
| *** pelix has quit IRC | 17:40 | |
| clarkb | beep beep beep | 17:40 |
| clarkb | pause | 17:40 |
| clarkb | beep beep beep | 17:40 |
| jeblair | yolanda: yes | 17:41 |
| yolanda | jeblair, ok, makes sense for me now | 17:42 |
| anteaya | krtaylor: https://wiki.openstack.org/w/index.php?title=Meetings/ThirdParty&diff=prev&oldid=57015 | 17:42 |
| yolanda | i'll work on it this week | 17:42 |
| anteaya | krtaylor: I had to create a new agenda for today and add a date to last week's agenda | 17:42 |
| anteaya | krtaylor: you can start the meeting and then use #chair to add me as co-chair | 17:43 |
| anteaya | krtaylor: okay for you? | 17:43 |
| jeblair | yolanda: i left a specific suggestion inline about how to do the build once, and pass the result to N-number of upload threads | 17:43 |
| *** rdopieralski has joined #openstack-infra | 17:43 | |
| krtaylor | anteaya, sure, will do | 17:43 |
| anteaya | krtaylor: also I haven't seen ilya from stackalytics add an agenda item, but do ask if anyone is there to discuss the driverlog page (we haven't finished that discussion) or the new ci status page (that was presented on the ml) | 17:44 |
| anteaya | krtaylor: thanks, and let me know if you don't know what I am talking about, I'm referecing the ml threads from last night | 17:44 |
| *** Sukhdev has quit IRC | 17:45 | |
| *** rcarrillocruz has joined #openstack-infra | 17:45 | |
| *** rcarrill` has quit IRC | 17:45 | |
| yolanda | jeblair, i saw it, thx | 17:46 |
| fungi | okay, tests all came back successful on the updated https://review.openstack.org/#/q/status:open+project:openstack-infra/config+branch:master+topic:normalize-acls,n,z patch series. some were seen as trivial rebases, others picked up a few additions | 17:46 |
| *** otherwiseguy has quit IRC | 17:46 | |
| jgriffith | jeblair: you don't by chance have any insight to why py26 might not be able to generate conf options correctly versus py27? | 17:46 |
| jgriffith | jeblair: https://review.openstack.org/#/c/103426/ | 17:46 |
| jgriffith | jeblair: result of the recent oslo messaging update | 17:47 |
| ryanpetrello | anybody mind taking a peek at https://review.openstack.org/#/c/103146/1 ? I’d like to get this in to release a new version of pecan soon | 17:48 |
| mordred | ryanpetrello: krotscheck and NikitaKonovalov were asking something about pecan a few days ago - I think specifically about supporting return codes or something ... | 17:49 |
| mordred | krotscheck, NikitaKonovalov: if you haven't met hi, ryanpetrello is the pecan guy | 17:49 |
| ryanpetrello | mordred: related to a new release, or just a pecan question in general? | 17:49 |
| *** rcarrillocruz has quit IRC | 17:49 | |
| *** rcarrillocruz has joined #openstack-infra | 17:50 | |
| mordred | ryanpetrello: I think it was related to pecan in general - and I seem to remember something that we thought might make a good upstream patch - but it was more than 1 hour ago, so I have forgotten | 17:50 |
| clarkb | fungi: so you are saying we should review that stack really quickly now so that we can stop rebasing it? | 17:50 |
| ryanpetrello | mordred: heh, yea, feel free to send folks to #pecanpy w/ questions | 17:51 |
| fungi | jgriffith: i wonder if that config generation is at all dynamic based on support it finds on the system (maybe redis is installed on our py27 test platform but not for py26 for example)? | 17:51 |
| *** otherwiseguy has joined #openstack-infra | 17:51 | |
| *** dangers is now known as dangers_away | 17:51 | |
| mordred | ryanpetrello: +A'd the patch | 17:51 |
| mordred | clarkb: yes. | 17:51 |
| ryanpetrello | (thanks) | 17:51 |
| jgriffith | fungi: oh... interesting | 17:51 |
| fungi | clarkb: jeblair and mordred went through it moments ago | 17:51 |
| fungi | jgriffith: that was only a wild guess | 17:51 |
| krtaylor | anteaya, were you saying you wanted me to chair today? | 17:51 |
| mordred | I'm +A on the whole stack | 17:51 |
| clarkb | fungi: cool will poke at it now too | 17:51 |
| jgriffith | fungi: it's dynamic based on code inspection though I thought | 17:51 |
| fungi | jgriffith: i don't know much about the config generator | 17:51 |
| *** dangers_away has quit IRC | 17:51 | |
| anteaya | krtaylor: yes, I am saying I would like you to chair today | 17:51 |
| clarkb | mordred: also, I suppose I should wait to see what rax says before I do anything new with volumes | 17:51 |
| jgriffith | fungi: yeah... it scrubs through the FLAGS using some OSLO tools | 17:52 |
| anteaya | krtaylor: start the meeting then add me as chair | 17:52 |
| jgriffith | fungi: I've asked on OSLO channel but crickets | 17:52 |
| *** devkulkarni has joined #openstack-infra | 17:52 | |
| jgriffith | fungi: so you guys are my go to (as much as I'm sure it pains you) for things like this | 17:52 |
| krtaylor | anteaya, ok, we were discussing next week too :) | 17:52 |
| anteaya | krtaylor: you can learn how to drive and I'll be around today | 17:52 |
| fungi | jgriffith: mainly just grasping at straws for what could cause a config section to be generated in the 27 job and not in the 26 job | 17:52 |
| anteaya | yes today is practice | 17:52 |
| krtaylor | anteaya, sure, np, I'll just need practice with the hash commands | 17:52 |
| anteaya | krtaylor: when I am still around, so you can ask questions | 17:52 |
| jgriffith | regardless, it appears that updating the RPC lib has some cost that wasn't expected | 17:52 |
| anteaya | krtaylor: great | 17:52 |
| anteaya | krtaylor: do your best and ask as we go | 17:53 |
| jgriffith | fungi: fair enough... I'm going to keep hollaring in all channels and ML *again* | 17:53 |
| anteaya | seeing someone ask and learn would be good for the group | 17:53 |
| *** pafuent has left #openstack-infra | 17:53 | |
| krtaylor | anteaya, I have run meetings before in other opensource projects, I'm just a bit rusty | 17:53 |
| jgriffith | fungi: thanks for ack'ing | 17:53 |
| fungi | np | 17:53 |
| anteaya | krtaylor: today is the day to knock off the rust | 17:53 |
| krtaylor | anteaya, no better time than the present | 17:53 |
| anteaya | exactly | 17:55 |
| *** rcarrillocruz has quit IRC | 17:56 | |
| mordred | clarkb: ok. I've +2'd/+A'd some of them and +2'd the rest - | 17:56 |
| openstackgerrit | A change was merged to openstack-infra/config: Switch jobs to use the build-timeout macro https://review.openstack.org/95933 | 17:57 |
| *** dangers_away has joined #openstack-infra | 17:58 | |
| *** dangers_away is now known as dangers | 17:59 | |
| *** thedodd has quit IRC | 17:59 | |
| *** ilyashakhat__ has joined #openstack-infra | 17:59 | |
| *** ryanpetrello has quit IRC | 18:00 | |
| openstackgerrit | A change was merged to openstack-infra/config: Add a script to normalize Gerrit ACLs https://review.openstack.org/93065 | 18:00 |
| *** rcarrillocruz has joined #openstack-infra | 18:01 | |
| anteaya | krtaylor: you're up | 18:01 |
| *** Ryan_Lane has joined #openstack-infra | 18:01 | |
| jeblair | clarkb: are you reviewing fungi's stack, or should i just +A? | 18:02 |
| openstackgerrit | A change was merged to openstack-infra/config: Normalize Gerrit ACL documentation https://review.openstack.org/93064 | 18:02 |
| clarkb | jeblair: I am reviewing it | 18:03 |
| clarkb | jeblair: but feel free to +A if you have beaten me to it | 18:03 |
| *** dkehn__ is now known as dkehnx | 18:03 | |
| jeblair | clarkb: the A is yours | 18:03 |
| clarkb | (I was trying to skim the 5k line change to make sure nothing funny happened) | 18:03 |
| openstackgerrit | A change was merged to openstack-infra/config: Gate pecan against designate. https://review.openstack.org/103146 | 18:03 |
| clarkb | which may be a bit silly | 18:03 |
| openstackgerrit | Andreas Jaeger proposed a change to openstack-infra/config: IRC setup for security-doc and training-guides https://review.openstack.org/103597 | 18:03 |
| *** wenlock has quit IRC | 18:04 | |
| *** ryanpetrello has joined #openstack-infra | 18:06 | |
| fungi | clarkb: not silly at all. i spent a ton of time staring at that one too | 18:08 |
| *** rcarrillocruz has quit IRC | 18:08 | |
| *** sarob has quit IRC | 18:09 | |
| *** arnaud_ has quit IRC | 18:11 | |
| *** salv-orlando has joined #openstack-infra | 18:11 | |
| *** andreykurilin has quit IRC | 18:13 | |
| *** salv-orlando has quit IRC | 18:14 | |
| ryanpetrello | clarkb: I’m seeing the same “NOT_REGISTERED” issue w/ the new pecan-designate tests as I saw last week w/ the ceilometer ones | 18:17 |
| *** SumitNaiksatam has joined #openstack-infra | 18:18 | |
| ryanpetrello | e.g., https://review.openstack.org/#/c/103611/ | 18:18 |
| *** andreaf_ has joined #openstack-infra | 18:20 | |
| fungi | ryanpetrello: i'm not finding those jobs as existing on any of our jenkins masters. i'll try manually rerunning a full jjb update on one and see what happens | 18:21 |
| ryanpetrello | fungi: okay, thanks | 18:21 |
| fungi | ryanpetrello: do you have the link to the change which added those jobs handy? | 18:21 |
| ryanpetrello | yep, sec | 18:22 |
| ryanpetrello | fungi: https://review.openstack.org/#/c/103146/ | 18:22 |
| *** andreaf_ has quit IRC | 18:23 | |
| fungi | ryanpetrello: oh! this merged 20 minutes ago | 18:23 |
| *** cnesa has quit IRC | 18:23 | |
| fungi | it may be that zuul picked up the update before jjb has finished running on any of our jenkins masters | 18:23 |
| ryanpetrello | ah, gotcha | 18:23 |
| ryanpetrello | okay, sounds good, I’ll give it an hour or so | 18:23 |
| ryanpetrello | I was just seeing this same error w/ the ceilometer stuff last week when clark moved them explicitly to trusty nodes | 18:24 |
| *** andreaf_ has joined #openstack-infra | 18:24 | |
| jeblair | ryanpetrello: if you want to be safer in the future, you can make 2 changes -- add job to jjb, then add it to zuul | 18:24 |
| ryanpetrello | okay, gotcha | 18:24 |
| mordred | fungi: I wonder if we should start running puppet on the jenkins masters before we run on them zuul? | 18:24 |
| jeblair | ryanpetrello: one change or two mostly depends on your tolerance for this kind of race condition :) | 18:24 |
| *** jp_at_hp has quit IRC | 18:24 | |
| mordred | fungi: we have the power to do that now | 18:24 |
| mordred | jeblair: ^^ | 18:24 |
| *** thedodd has joined #openstack-infra | 18:25 | |
| mordred | so, we could do "git-farm, gerrit, jenkins farm, zuul, everything else" | 18:25 |
| jeblair | mordred: wfm; btw, any chance we can do that scatter-gather? | 18:25 |
| *** arnaud_ has joined #openstack-infra | 18:25 | |
| mordred | jeblair: can you say a few more words just so I know what you're thinking? I think we probably can ... | 18:26 |
| *** andreaf_ has quit IRC | 18:26 | |
| *** om has quit IRC | 18:26 | |
| mordred | zomg. mirror move ALMOST done | 18:26 |
| jeblair | mordred: right now git-farm is sequential, but we really could run all 5 in parallel and then iff all 5 succeed, continue, else, fail | 18:26 |
| openstackgerrit | A change was merged to openstack-infra/config: Strip/condense whitespace and sort Gerrit ACLs https://review.openstack.org/93066 | 18:26 |
| mordred | jeblair: in the new ansible world order, the git-farm is parallel | 18:27 |
| jeblair | mordred: same applies to jenkins masters/zuul | 18:27 |
| *** rcarrillocruz has joined #openstack-infra | 18:27 | |
| *** andreaf_ has joined #openstack-infra | 18:27 | |
| mordred | jeblair: so, YES! | 18:27 |
| jeblair | good, then let's do that with ansible rather than the current system, so we don't slow it down too much | 18:27 |
| mordred | jeblair: btw - I rebased and squashed that change and did not apply it over the weekend like I'd planned | 18:28 |
| mordred | jeblair: as I tested it extensively and then tweaked some | 18:28 |
| mordred | jeblair: so patch one needs a re-review - but also has been tested | 18:28 |
| *** hashar has joined #openstack-infra | 18:28 | |
| mriedem | anyone here know anything about apache mesos? | 18:29 |
| openstackgerrit | A change was merged to openstack-infra/config: Get rid of unneeded tags create in Gerrit ACLs https://review.openstack.org/93067 | 18:29 |
| mordred | mriedem: I think I heard the name at one point | 18:31 |
| mriedem | mordred: yeah trying to figure out the difference between mesos and openstack, or if they play together or what, read an article from ebay on how they ran mesos on top of openstack with jenkins | 18:32 |
| *** lttrl has joined #openstack-infra | 18:32 | |
| mriedem | i'm getting buzzword overkill since docker is also in the mix | 18:32 |
| mordred | jeblair: thanks for the review - I'll babysit that in next time I've got spare minutes | 18:32 |
| mordred | mriedem: god | 18:33 |
| clarkb | fungi: jeblair mordred ok fungi's stack is compeltely approved now | 18:33 |
| openstackgerrit | Ken Giusti proposed a change to openstack/requirements: Add pyngus for oslo.messaging https://review.openstack.org/103616 | 18:33 |
| clarkb | mriedem: mesos is just a job scheduler thing | 18:33 |
| mriedem | yeah sounds like orchestration | 18:33 |
| mordred | ah. seems like zuul+nodepool to me | 18:33 |
| clarkb | mriedem: so you tell it I have a foo, and it needs this sort of cpu time make it go | 18:33 |
| clarkb | mordred: less specific but yesish | 18:33 |
| clarkb | mordred: I actually think nodepool could add mesos nodes nad zuul schedule with mesos instead of gearman | 18:34 |
| clarkb | because mesos doesn't really do smarts about I need more nodes or what it means to test git trees | 18:34 |
| mordred | oh. weird. ok | 18:34 |
| clarkb | you can however implement custom schedulers for it that could do nodepool and zuul | 18:34 |
| *** Sukhdev has joined #openstack-infra | 18:35 | |
| hashar | clarkb: oh please stop changing your stack all the time. I have trouble keeping up :-D | 18:35 |
| clarkb | hashar: I am not suggesting we change :) | 18:35 |
| openstackgerrit | A change was merged to openstack-infra/config: Remove active project.state in Gerrit ACLs https://review.openstack.org/93068 | 18:35 |
| clarkb | just trying to describe what mesos does wtih examples | 18:35 |
| hashar | =-) | 18:36 |
| clarkb | mordred: the problem space I actually think mesos might be kind of neat for is in baremetal land. Ironic spins up a datacenter for you then you throw lots of mesos work at it | 18:36 |
| clarkb | mordred: because in theory you can restrict the crazyness in that case and have relatively secure bare metal workloads | 18:37 |
| openstackgerrit | A change was merged to openstack-infra/config: Strip *.owner Administrators from Gerrit ACLs https://review.openstack.org/93069 | 18:37 |
| openstackgerrit | A change was merged to openstack-infra/config: Sort exclusiveGroupPermissions in Gerrit ACLs https://review.openstack.org/93070 | 18:38 |
| openstackgerrit | A change was merged to openstack-infra/config: Combine infra/ci core/admin gerrit groups https://review.openstack.org/102271 | 18:38 |
| clarkb | my held bare-trusty node appears to have been cleaned up /me grabs another | 18:40 |
| fungi | clarkb: oh, i think i might have overzealously deleted really old nodes when cleaning up from the jenkins03 problem this morning | 18:41 |
| *** dtantsur has quit IRC | 18:41 | |
| fungi | sorry about that | 18:41 |
| *** bdpayne has joined #openstack-infra | 18:42 | |
| clarkb | fungi: thats ok | 18:42 |
| clarkb | there wasn't anything important on the node | 18:42 |
| clarkb | just having a node is what I nee | 18:42 |
| fungi | okay, cool | 18:42 |
| *** dtantsur has joined #openstack-infra | 18:43 | |
| clarkb | going to test oslo on trusty now | 18:43 |
| clarkb | and start looking at projects like designate and trove | 18:43 |
| bdpayne | OSSG has been using the #openstack-security channel for some time now. I'm trying to register #openstack-security with chanserv so that the infra team has access. However, I'm seeing an error "#openstack-security is not registered". Thoughts? | 18:43 |
| mordred | if you've been in there non-registered and lost ops on teh channel, we may have to get jeblair to use his super cow powers | 18:44 |
| Ajaeger | bdpayne: you're quick - thanks ;) | 18:45 |
| bdpayne | :-) | 18:45 |
| mordred | fungi, clarkb: the mirror file move is complete, I'd like to land https://review.openstack.org/#/c/103353/ | 18:45 |
| bdpayne | let me check, someone else may have ops in that channel at this point | 18:46 |
| mordred | bdpayne: if not, I Think jeblair is on a call right now, but he has the power to help | 18:46 |
| mordred | fungi, clarkb: I've disabled puppet on static.o.o | 18:46 |
| bdpayne | ok, I will need his help... doesn't look like anyone online has ops | 18:46 |
| SlickNik | clarkb: FWIW, I'm attempting to run trove tests in devstack-gate on a trusty image atm. | 18:46 |
| clarkb | mordred: mirror file move? | 18:46 |
| clarkb | SlickNik: well devstack-gate is already trusty :) | 18:46 |
| mordred | clarkb: I built the bandersnatch mirror on teh old volume, needed to move it to the new volume | 18:47 |
| bdpayne | jeblair please ping me when you have a sec... thanks :-) | 18:47 |
| clarkb | SlickNik: we moved all of the integration tests at once becuase they are integration tests | 18:47 |
| mordred | that move took a couple of hours | 18:47 |
| fungi | bdpayne: one alternative option is if you can convince everyone to part the channel, then you can join as a chanop | 18:48 |
| clarkb | mordred: are the old jobs being turned off? | 18:48 |
| clarkb | mordred: it looks like you are keeping the jenkins jobs | 18:48 |
| SlickNik | clarkb: ah, nice. I brought this one up myself (not using nodepool) to test moving the trove functional tests to a devstack-gate setup. | 18:48 |
| *** yolanda has quit IRC | 18:48 | |
| mordred | clarkb: yeah - there isa sequence - this is purely about adding the new mirror and moving the directory | 18:49 |
| mordred | clarkb: so we should be able to test things against the new mirror before removing anything that currently works | 18:49 |
| clarkb | mordred: gotcha | 18:49 |
| clarkb | mordred: because we have the openstack/ prefix | 18:49 |
| mordred | yup | 18:51 |
| *** andreaf has quit IRC | 18:51 | |
| clarkb | lgtm | 18:51 |
| clarkb | (the second change in that stack helped me grok all that) | 18:51 |
| clarkb | I didn't approve anything as it sounded like you are babysitting | 18:52 |
| *** andreaf has joined #openstack-infra | 18:52 | |
| fungi | mordred: i'm cool with it, though wondering whether we should consider latest instead of present for the bandersnatch package | 18:52 |
| mordred | fungi: we probably do - but I figured while we learn a little more about the stability, we know this one works :) | 18:53 |
| fungi | sure, wfm | 18:53 |
| *** otherwiseguy has quit IRC | 18:53 | |
| fungi | ugh. i think something is still going on with jenkins03 even after the reboot. it keeps accumulating "ready" nodes at a fairly rapid clip which don't actually pick up jobs, and are starving us out of rarer node types as a result | 18:54 |
| fungi | right now it has >80 nodes in a "ready" state while its peers have <10 | 18:55 |
| *** devkulkarni has quit IRC | 18:55 | |
| *** devkulkarni has joined #openstack-infra | 18:56 | |
| *** joa has quit IRC | 18:56 | |
| *** rcarrill` has joined #openstack-infra | 18:56 | |
| *** joa has joined #openstack-infra | 18:57 | |
| fungi | 70 of those are 1-3 hours old (no older than that because i deleted any which were ready more than an hour on it a couple hours ago) | 18:57 |
| *** mrda-away is now known as mrda | 18:58 | |
| *** rcarrillocruz has quit IRC | 18:58 | |
| fungi | oh wow, they're running jobs | 18:59 |
| clarkb | fungi: :/ | 18:59 |
| fungi | it's not reporting them back to nodepool as being in use | 18:59 |
| clarkb | fungi: sounds like nodepool may not be getting events | 18:59 |
| clarkb | ya | 19:00 |
| clarkb | probably need to restart nodepool | 19:00 |
| clarkb | and logstash jenkins/gearman client. I can do logstash | 19:00 |
| fungi | i bet the network derp jenkins03 had this morning left a hung established socket from nodepoold to the jenkins api endpoint | 19:00 |
| fungi | or something along those lines | 19:00 |
| anteaya | so as an update, I couldn't get stackalytics to agree to stop making up definitions of success for third party ci systems on their pages | 19:01 |
| clarkb | logstash side is done | 19:01 |
| anteaya | so I will have to keep following them around on the ml | 19:02 |
| anteaya | they seem determined to want to define this for openstack | 19:02 |
| clarkb | dhellmann: are you around? I am looking at the oslotest cross project tests to test that before switching them to trusty | 19:02 |
| krtaylor | anteaya, do you have time to put some thoughts down around definitions? | 19:02 |
| anteaya | rather than contribute and allow us the time and space to define this for ourselves | 19:02 |
| anteaya | krtaylor: yes | 19:02 |
| anteaya | ttx ^^ | 19:02 |
| clarkb | dhellmann: and well I am not groking why they run in devstack | 19:03 |
| openstackgerrit | A change was merged to openstack-infra/config: Start building a bandersnatch mirror https://review.openstack.org/103353 | 19:03 |
| clarkb | dhellmann: is there a reason for devstack? | 19:03 |
| krtaylor | anteaya, I can bring it up again next week, but maybe we can start framing a glossary of sorts | 19:03 |
| *** homeless has quit IRC | 19:03 | |
| anteaya | krtaylor: I think a glossary is a great idea and will be an ongoing project | 19:03 |
| krtaylor | anteaya, ++, I'll get it started this week | 19:04 |
| anteaya | krtaylor: thank you | 19:04 |
| * Ajaeger just reads glossary without context - and thus wants to point out http://docs.openstack.org/glossary/ | 19:04 | |
| mordred | boom. new mirror live | 19:05 |
| *** homeless has joined #openstack-infra | 19:05 | |
| mordred | clarkb, fungi: ^^ | 19:05 |
| mordred | http://pypi.openstack.org/simple can now be used | 19:05 |
| clarkb | cool | 19:05 |
| *** annegent_ has joined #openstack-infra | 19:05 | |
| krtaylor | Ajaeger, thanks, this is for 3rd party CI specifically, but I'll make sure it is consistent or uses any existing definitions | 19:06 |
| enikanorov__ | anteaya: could you please explain your confusion around the term 'success'? | 19:06 |
| openstackgerrit | Spencer Krum proposed a change to openstack-infra/config: Add node def for puppet3 master https://review.openstack.org/83678 | 19:07 |
| dstufft | mordred: tls imo | 19:07 |
| *** adam_g` is now known as adam_g | 19:07 | |
| *** dims has quit IRC | 19:08 | |
| mordred | dstufft: aroo? | 19:08 |
| dstufft | mordred: https:// | 19:08 |
| dstufft | :D | 19:08 |
| anteaya | enikanorov__: hi | 19:09 |
| Ajaeger | krtaylor: I pointed out the glossary for the documentation project, I guess it's unrelated but if you can use it, great... | 19:09 |
| enikanorov__ | hi anteaya | 19:09 |
| mordred | dstufft: yah. on the tdl | 19:09 |
| mordred | dstufft: :) | 19:09 |
| dstufft | mordred: yay | 19:09 |
| anteaya | enikanorov__: can you share the context for your question? have you been following all the email replies and the third party meeting? | 19:09 |
| anteaya | or just party? | 19:10 |
| anteaya | part | 19:10 |
| fungi | mordred: the https://review.openstack.org/103361 change to remove the mirror slave definitions from the global site manifest needs a rebase still | 19:10 |
| krotscheck | So I’ve been thinking. There’s a lot of vagueness about what StoryBoard “Is going to be”. I kindof feel that having a long-term vision doc would be super useful, because it’ll put our requirements and features in context. Something… higher level, that doesn’t make statements about implementation, but describes what its raison d’etre is. We already have part of that in the “About” page, but would it make sense to actually hav | 19:10 |
| krotscheck | vision spec in infra-specs? | 19:10 |
| *** james_li_ has joined #openstack-infra | 19:10 | |
| *** msabramo has quit IRC | 19:10 | |
| clarkb | krotscheck: your first long message got cut off at sense to actually hav | 19:10 |
| krotscheck | ...., but would it make sense to actually have a vision spec in infra-specs? | 19:11 |
| enikanorov__ | anteaya: i'm asking in the context of email thread | 19:11 |
| krotscheck | Client fail | 19:11 |
| fungi | dstufft: though also a purported reason for not doing https with our pypi mirror is to discourage people for using it for anything important (so they'll not be tempted to use it for their own production systems instead of the real pypi, as an example) | 19:11 |
| *** homeless has quit IRC | 19:11 | |
| anteaya | enikanorov__: ah so you didn't attend the third party meeting, okay | 19:11 |
| *** nati_ueno has quit IRC | 19:11 | |
| *** james_li has quit IRC | 19:12 | |
| *** james_li_ is now known as james_li | 19:12 | |
| *** homeless has joined #openstack-infra | 19:12 | |
| *** nati_ueno has joined #openstack-infra | 19:12 | |
| dstufft | fungi: that'd be good except I bet people will still do it and just be insecure ;( | 19:12 |
| *** arnaud_ has quit IRC | 19:12 | |
| anteaya | enikanorov__: so in the third party meeting, we discussed that the use of the term success in reference to evaluating a third party ci system is a misnomer | 19:12 |
| enikanorov__ | testing results expressed as a vote on gerrit always seemed something natural to me, thought without any special meaning. "if it's green - skip it, if it's red - look at it" | 19:12 |
| clarkb | krotscheck: I don't think it will hurt, but also worry that decoupling that broad vision from the project itself may hurt | 19:12 |
| fungi | dstufft: probably so | 19:12 |
| anteaya | since returning 100% build status successful is not indicative of a trustworthy system | 19:12 |
| enikanorov__ | that's true | 19:12 |
| anteaya | right | 19:13 |
| krotscheck | clarkb: To me that says it should live in the project documentation… | 19:13 |
| clarkb | krotscheck: ya I think that isn't a terrible place for it | 19:13 |
| enikanorov__ | so I'm trying to understand, is there an issue with stackalytics or with most of 3rd part CIs | 19:13 |
| anteaya | so until we have some form of agreed upon algorithm, from the community, to decide how to assess fitness of a third party system, using the term success if confusing | 19:13 |
| *** markmcclain has joined #openstack-infra | 19:14 | |
| anteaya | the issue with stackalytics is the use of the word success | 19:14 |
| anteaya | in regards to third party systems | 19:14 |
| anteaya | it is premature | 19:14 |
| anteaya | we don't have an algorithm to asssess fitness | 19:14 |
| sdague | clarkb: the grenade log fix definitely seems to help. Now when we get rediculous spikes in ES, we at least catch up again | 19:14 |
| anteaya | and every time stackalytics uses success, the media thinks it has a meaning it doesn't have | 19:14 |
| anteaya | so please stop | 19:15 |
| anteaya | and join us in the conversation to develop an algorighm to assess fitness of third party ci systems | 19:15 |
| sdague | clarkb: I had a question on indexing idea if you have a few minutes to kick something around | 19:15 |
| anteaya | since having to chase people down on the ml is a waste of my energy | 19:15 |
| anteaya | I would rather participate in the discussion of the algorithm to assess fitness of third party ci systems | 19:16 |
| anteaya | and then everyone can use that algorithm | 19:16 |
| anteaya | including stackalytics | 19:16 |
| enikanorov__ | So are you suggesting just to be more descriptive explaining what 'Success, %' means? | 19:16 |
| anteaya | I'm suggesting you take a step back | 19:16 |
| anteaya | and particpate in the discussion | 19:16 |
| anteaya | and acknowledge taht noone knows what it means yet | 19:17 |
| *** homeless has quit IRC | 19:17 | |
| anteaya | and we have to arrive at the meaning together | 19:17 |
| *** pballand has joined #openstack-infra | 19:17 | |
| enikanorov__ | well, beside the discussion one may want to know bare numbers | 19:17 |
| anteaya | I should hope so | 19:17 |
| anteaya | and number should come out in the discussion | 19:17 |
| enikanorov__ | so... what would you call the number of +1s CI has ever voted? | 19:18 |
| anteaya | but please participate in the discussion | 19:18 |
| anteaya | please stop trying to fragemtn my focus | 19:18 |
| *** bnemec has quit IRC | 19:18 | |
| anteaya | I would like to focus on the algorithm to assess fitness for third party ci systems | 19:18 |
| anteaya | please participate or don't | 19:18 |
| enikanorov__ | that seems to be a discussion about the terms. That raraly happens to be productive | 19:18 |
| enikanorov__ | if you have better suggestion for the metric name - that's fine | 19:19 |
| anteaya | but if you choose not to, please stop trying to put definitons in this space | 19:19 |
| enikanorov__ | the funny thing is that we don't put definitions | 19:19 |
| anteaya | yes you are | 19:19 |
| anteaya | you just think you aren't | 19:19 |
| enikanorov__ | the number is not an interpretation | 19:19 |
| anteaya | that is the problem | 19:19 |
| enikanorov__ | it's a mere metric, simple count | 19:19 |
| anteaya | you aren't paying attention to houw your information is consumed | 19:19 |
| anteaya | no it isnt' | 19:20 |
| anteaya | the count is 1 | 19:20 |
| enikanorov__ | how it is consumed IS the interpretation | 19:20 |
| *** homeless_ has joined #openstack-infra | 19:20 | |
| anteaya | enikanorov__: are you willing to participate with the rest of the community on this | 19:20 |
| enikanorov__ | interpretation deserves discussion, terms - don't | 19:20 |
| anteaya | yes they do | 19:20 |
| *** rcarrill` has quit IRC | 19:20 | |
| *** bnemec has joined #openstack-infra | 19:21 | |
| anteaya | enikanorov__: are you willing to participate with the rest of the community on this? | 19:21 |
| nibalizer | jeblair: thanks for the +2 on the puppet3 master review :) | 19:22 |
| enikanorov__ | anteaya: that's a discussion of what name to put on top of the +1 rate column | 19:22 |
| anteaya | enikanorov__: no, it is the willingness to participate with the communtiy | 19:23 |
| enikanorov__ | is that what you want to discuss? the number in the column will not change, not will change the simple algorithm that collects numbers | 19:23 |
| anteaya | as teh community defines the algorthim to decide how to assess the fitness of third party ci systems | 19:23 |
| enikanorov__ | anteaya: i'm willing to participate in productive things | 19:23 |
| anteaya | hwo to communicate | 19:23 |
| anteaya | well this is necessary | 19:23 |
| anteaya | third party ci systems affect a lot of people | 19:24 |
| enikanorov__ | yes, they do | 19:24 |
| anteaya | and people need to be given accurate information | 19:25 |
| anteaya | and they aren't right now | 19:25 |
| anteaya | and that is a problem that needs to be addressed | 19:25 |
| enikanorov__ | the only issue i may see here is more verbose about what 'Success, %' means | 19:25 |
| clarkb | sdague: after lunch I can | 19:25 |
| anteaya | are you willing to particpate with the communtiy in ensuring accurate information about third party systems is conveyed? | 19:25 |
| sdague | clarkb: sounds good, ping me when you get back | 19:25 |
| *** rcarrillocruz has joined #openstack-infra | 19:26 | |
| anteaya | enikanorov__: then you aren't understanding the concern | 19:26 |
| anteaya | and I dno't know why | 19:26 |
| anteaya | you are otherwise an intelligent person | 19:26 |
| enikanorov__ | because the concern doesn't deserve the discussion | 19:26 |
| anteaya | yes it does | 19:26 |
| enikanorov__ | 1+1=2 doesn't deserve the discussion | 19:26 |
| anteaya | yes it does | 19:26 |
| anteaya | this needs to be discussed | 19:27 |
| anteaya | developers are confused | 19:27 |
| anteaya | and you are giving out inaccurate information on your pages | 19:27 |
| enikanorov__ | what are their confusion? | 19:27 |
| anteaya | I just told you | 19:27 |
| anteaya | and you dismissed it | 19:27 |
| *** otherwiseguy has joined #openstack-infra | 19:27 | |
| enikanorov__ | no, i don't dismiss it, sorry. I'm telling you that stackalytics provides very simple basic info | 19:27 |
| enikanorov__ | and the interpretation is responsibility of those who consume it | 19:28 |
| enikanorov__ | even more, it's just an aggregate of what gerrit has | 19:28 |
| fungi | mordred: i must say that i find 103355 extremely exciting. i hadn't given much though to how much "the one mirror" would simplify our job definitions and slave scripts | 19:28 |
| anteaya | is stackalytics willing to participate in teh definiton of the algorithm to assess fitness of third party ci systems? | 19:28 |
| jaypipes | anteaya: yes. | 19:29 |
| enikanorov__ | anteaya: stackalitycs is willing to improve based on reasonable suggestion of the community | 19:29 |
| anteaya | jaypipes: awesome | 19:29 |
| anteaya | jaypipes: thank you jaypipes | 19:29 |
| anteaya | jaypipes: I've missed you | 19:29 |
| *** eharney has quit IRC | 19:29 | |
| jaypipes | can we all calm down a bit, please? nobody has ill intent here. | 19:29 |
| enikanorov__ | that's true | 19:29 |
| *** julim has quit IRC | 19:30 | |
| jaypipes | let's discuss on the ML about an algorithm for assessing whether a driver is successfully passing a CI test suite. | 19:30 |
| jaypipes | I believe we all have the same goal of providing as much accurate information as possible. | 19:31 |
| jaypipes | let's make that a reality and discuss on the ML. I will write a new post to the ML offering my thoughts on such an algorithm, and let's take it from there, ok, anteaya? | 19:31 |
| bodepd | how are you guys connecting jenkins clients to the master? I didn't see any code that did it | 19:32 |
| anteaya | jaypipes: wonderful, thank you | 19:32 |
| jaypipes | bodepd: look in nodepool-scripts... one sec, grabbing a link for you. | 19:32 |
| bodepd | jaypipes: thanks! | 19:33 |
| *** julim has joined #openstack-infra | 19:33 | |
| fungi | bodepd: are you asking how nodepool registers a new slave with a jenkins master? | 19:33 |
| jaypipes | bodepd: start here :) https://github.com/openstack-infra/config/blob/master/modules/openstack_project/files/nodepool/scripts/prepare_node.sh | 19:33 |
| bodepd | fungi: yes | 19:34 |
| *** sileht_ has quit IRC | 19:34 | |
| fungi | bodepd: if so, that's going to be more in the source for https://git.openstack.org/cgit/openstack-infra/nodepool/tree/nodepool | 19:34 |
| *** sileht has joined #openstack-infra | 19:34 | |
| fungi | bodepd: but the short answer is that it uses the python-jenkins library to make the api calls | 19:35 |
| fungi | bodepd: https://git.openstack.org/cgit/stackforge/python-jenkins/tree/ | 19:35 |
| *** annegent_ has quit IRC | 19:35 | |
| *** amitpp has joined #openstack-infra | 19:35 | |
| openstackgerrit | Antoine Musso proposed a change to openstack-infra/zuul: Sphinx doc for zuul {promote,show} https://review.openstack.org/103639 | 19:36 |
| *** reed has quit IRC | 19:36 | |
| *** reed_ is now known as reed | 19:36 | |
| *** reed has quit IRC | 19:36 | |
| *** reed has joined #openstack-infra | 19:36 | |
| enikanorov__ | jaypipes: looking forward for such algorithm... | 19:36 |
| jaypipes | enikanorov_, anteaya: got a link to that stackalytics report page? | 19:37 |
| fungi | hashar: awesome! i was just playing around with the zuul show cli subcommand earlier today | 19:37 |
| enikanorov__ | jaypipes: for example: http://stackalytics.com/report/ci/neutron/7 | 19:38 |
| jaypipes | thx! | 19:38 |
| enikanorov__ | (we initially were discussing 3rd part Cis, not vendor drivers though) | 19:38 |
| hashar | fungi: I got it installed last week (finally upgraded Zuul on my instal) and looking at http://ci.openstack.org/zuul/client.html I was wondering what can be done | 19:38 |
| hashar | fungi: turns out my PrettyTable version is too old :-( | 19:39 |
| jaypipes | enikanorov_: yes, understood. | 19:39 |
| jaypipes | hashar: PrettyOldTable? :P | 19:39 |
| *** annegent_ has joined #openstack-infra | 19:39 | |
| hashar | jaypipes: yeah we run Precise which has 0.5 and 0.6 has slightly different API :-d | 19:39 |
| *** annegent_ has quit IRC | 19:40 | |
| hashar | one day I will find a solution to ship whatever deps I need instead of the deb packages | 19:40 |
| bodepd | fungi: if you guys did move to heat, where would this logic exist? | 19:40 |
| *** annegent_ has joined #openstack-infra | 19:40 | |
| bodepd | fungi: would it be part of userdata? moved to Puppet, or a separete script that runs post provisionig? | 19:40 |
| bodepd | fungi: or would you "Those are all good questions" :) | 19:41 |
| fungi | bodepd: i am ashamed to say that i haven't though that far ahead yet | 19:42 |
| fungi | s/though/thought/ | 19:42 |
| bodepd | fungi: for context, I put togehter some heat templates to deploy the bits of infra that I wanted | 19:42 |
| bodepd | fungi: they do everything except connect slaves at the moment | 19:43 |
| fungi | bodepd: oh, awesome! do you have a cloud with heat service exposed, or are you using heat as a standalone orchestration engine? | 19:43 |
| bodepd | fungi: rackspace cloud :) | 19:43 |
| *** ilyashakhat__ has quit IRC | 19:43 | |
| anteaya | krtaylor: for next week's third party meeting, be sure to mention the system's listed at the top of this etherpad: https://etherpad.openstack.org/p/ZLp9Ow3tNq | 19:44 |
| anteaya | krtaylor: we need to ensure systems have lots of communication that they need to take action | 19:44 |
| *** wenlock has joined #openstack-infra | 19:44 | |
| bodepd | fungi: I guess I could have also probably used HP cloud :) | 19:44 |
| bodepd | the issue that I am running into is that userdata is pretty purpose towards bootstrapping Puppet and I was hoping to make something generic/reusable | 19:45 |
| *** maxbit_ has joined #openstack-infra | 19:46 | |
| krtaylor | anteaya, understood, will do | 19:46 |
| bodepd | It feels like the heat scripts becoming less reusable if I start adding custom post-puppet-actions | 19:46 |
| bodepd | although, I guess it's doable... | 19:46 |
| krtaylor | anteaya, thats a great reference | 19:46 |
| *** lttrl has quit IRC | 19:48 | |
| *** markmcclain has quit IRC | 19:49 | |
| *** chandan_kumar has quit IRC | 19:52 | |
| bodepd | I could sit down and write jenkins_node types that use those libs. Is that something you guys would be interested in? | 19:52 |
| *** bcrochet has quit IRC | 19:54 | |
| *** nati_ueno has quit IRC | 19:55 | |
| *** markmc has quit IRC | 19:55 | |
| anteaya | krtaylor: thanks, yes very helpful | 19:55 |
| *** nati_ueno has joined #openstack-infra | 19:55 | |
| *** arnaud has joined #openstack-infra | 19:55 | |
| *** dims has joined #openstack-infra | 19:56 | |
| *** bcrochet has joined #openstack-infra | 19:57 | |
| *** cindyo has joined #openstack-infra | 19:59 | |
| jeblair | bdpayne: hi there | 19:59 |
| bdpayne | hi! | 19:59 |
| bdpayne | did you see the situation in the backlogs? | 19:59 |
| jeblair | bdpayne: you need #openstack-security registered, yeah? | 20:00 |
| *** julim has quit IRC | 20:00 | |
| bdpayne | yeah | 20:00 |
| jeblair | bdpayne: any chance you can convince everyone in the channel currently to leave? | 20:00 |
| *** ociuhandu has quit IRC | 20:00 | |
| bdpayne | heh, we were just talking about that | 20:00 |
| bdpayne | I can try... but likely that they are not all present | 20:00 |
| bdpayne | I'll hope over there now and see what I can do | 20:00 |
| jeblair | bdpayne: it's okay | 20:00 |
| *** atiwari has joined #openstack-infra | 20:00 | |
| jeblair | bdpayne: if you think it's going to be difficult, i'll try to ask a freenode op | 20:01 |
| clarkb | dhellmann: is devstack used purely for cogating? | 20:01 |
| clarkb | dhellmann: trying to make sure I test this mostly properly | 20:01 |
| clarkb | sdague: ok now is good | 20:01 |
| *** julim has joined #openstack-infra | 20:02 | |
| bdpayne | jeblair let's see if it works | 20:02 |
| bdpayne | we'll know in a few | 20:02 |
| *** Woodie has joined #openstack-infra | 20:02 | |
| *** W00die has quit IRC | 20:03 | |
| *** msabramo has joined #openstack-infra | 20:04 | |
| sdague | clarkb: so I was thinking about other information that would be really good to just know about a run | 20:04 |
| sdague | and one of them that pops up is which tests failed, by name | 20:04 |
| openstackgerrit | Andreas Jaeger proposed a change to openstack-infra/config: Cleanup manual jobs https://review.openstack.org/103646 | 20:05 |
| sdague | and I was wondering if we could get those indexed into a tag namespace | 20:05 |
| *** eharney has joined #openstack-infra | 20:05 | |
| *** basha has quit IRC | 20:06 | |
| clarkb | sdague: something like failed_tests: [name1, name2, name3] for documents associated with the job? | 20:07 |
| sdague | yep | 20:07 |
| clarkb | the most direct way to do that would be to ask the subunit log | 20:07 |
| sdague | right | 20:07 |
| clarkb | but those are huge... | 20:07 |
| clarkb | it might get old fetching and loading those for each test that fails | 20:08 |
| *** homeless_ has quit IRC | 20:08 | |
| sdague | so could we put the query server side? | 20:08 |
| clarkb | maybe we can do a testr failing at the end of a failed job and just index the data more naturally | 20:08 |
| hashar | nibalizer: I noticed a talk about puppet 3. Do you have your manifests migrated already? There are slight differences | 20:08 |
| sdague | clarkb: sure, that's an option as well | 20:08 |
| fungi | bodepd: userdata can be an executable script too. for my personal systems i pass a fairly complex shell script which does a lot of post-build configuration. its main downside is that anything launched from userdata has to complete before nova boot will return, and you don't get redirection of that to your calling terminal | 20:09 |
| *** Guest9812 has quit IRC | 20:10 | |
| sdague | clarkb: however you think is most sane to get this info, I'm happy to help implement | 20:10 |
| fungi | ryanpetrello: your jenkins jobs seem to exist now, so the "not registered" situation should cease | 20:10 |
| clarkb | ok, let me think about it but I have a hunch that just generating the data we want in the job and indexing it naturally seems best | 20:10 |
| clarkb | it reducses special code | 20:10 |
| sdague | but staring at new bugs recently I realized a lot of these I can actually pattern match in my head based on fail pattern | 20:10 |
| *** yfried_ has quit IRC | 20:10 | |
| ryanpetrello | fungi: thanks :) | 20:11 |
| *** yfried_ has joined #openstack-infra | 20:11 | |
| sdague | and so failure + log strings would probably help mitigate the need for multiline for some of the more tricky ones | 20:11 |
| clarkb | ya | 20:11 |
| nibalizer | hashar: not yet | 20:13 |
| nibalizer | but first step is to spin up a p3 master so you can see the problems | 20:13 |
| *** amitpp has quit IRC | 20:14 | |
| *** sarob has joined #openstack-infra | 20:14 | |
| hashar | nibalizer: we (Wikimedia) completed the migration from p2.7 to p3 last week iirc. Two of our ops wrote a catalog compiler that takes a node + facts and compile with puppet 2.7 and 3.0 | 20:14 |
| jeblair | bdpayne: all set | 20:14 |
| bdpayne | thanks | 20:14 |
| nibalizer | sweeet | 20:15 |
| bdpayne | did you also do the registration with openstack infra? | 20:15 |
| nibalizer | got that tool up somewhere? | 20:15 |
| jeblair | bdpayne: you can give yourself (and others) op access by making a change to ./modules/accessbot/templates/accessbot.config.erb in openstack-infra/config | 20:15 |
| hashar | nibalizer: that really helped the migration. example output http://puppet-compiler.wmflabs.org/87/html/ | 20:15 |
| jeblair | bdpayne: (sorry that's not documented yet) | 20:15 |
| jeblair | bdpayne: sorry, i mean ./modules/openstack_project/files/accessbot/channels.yaml | 20:16 |
| bdpayne | got it, I'll head over there while it's on my mind | 20:16 |
| *** crc32 has joined #openstack-infra | 20:16 | |
| *** homeless has joined #openstack-infra | 20:16 | |
| nibalizer | hashar: oh thats slick | 20:16 |
| bdpayne | jeblair but otherwise we should be all set? | 20:16 |
| nibalizer | is the tool open? | 20:16 |
| *** sarob has quit IRC | 20:16 | |
| *** alexpilotti has quit IRC | 20:16 | |
| openstackgerrit | Sean Dague proposed a change to openstack-dev/hacking: remove H904 https://review.openstack.org/101701 | 20:16 |
| jeblair | bdpayne: oh, actually, just add it to the list, and then i'll bootstrap you as an op | 20:16 |
| openstackgerrit | Sean Dague proposed a change to openstack-dev/hacking: remove H305 because it's python version specific https://review.openstack.org/101499 | 20:16 |
| nibalizer | or is it just shell around puppet catalog or whatever? | 20:16 |
| jeblair | bdpayne: yeah, you can also add it to eavesdrop if you want | 20:16 |
| hashar | nibalizer: some nodes have difference (blue links). | 20:16 |
| hashar | nibalizer: source code is at: git clone https://gerrit.wikimedia.org/r/p/operations/software.git && cd compare-puppet-catalogs | 20:17 |
| *** sarob has joined #openstack-infra | 20:17 | |
| hashar | nibalizer: author is Giuseppe Lavagetto he is europe based, feel free to contact me on my behalf. | 20:17 |
| *** james_li has quit IRC | 20:18 | |
| Ajaeger | bdpayne: once everything is setup for #openstack-security, please comment on https://review.openstack.org/103597 so that it gets reviewed again ;) | 20:18 |
| bdpayne | can do | 20:19 |
| nibalizer | hashar: awesome | 20:19 |
| nibalizer | thanks! | 20:19 |
| hashar | nibalizer: and the very basic Jenkins job that wrap around it http://paste.openstack.org/show/85180/ | 20:21 |
| anteaya | storms are overhead and I'm shutting off power | 20:21 |
| anteaya | back later | 20:21 |
| hashar | nibalizer: there is some other hacked stuff that are not in the conf though. But the compare-puppet-catalogs might well have everything you need. | 20:21 |
| *** dstanek_zzz is now known as dstanek | 20:22 | |
| openstackgerrit | Andreas Jaeger proposed a change to openstack-infra/config: Cleanup manual jobs https://review.openstack.org/103646 | 20:22 |
| jaypipes | anteaya: k, mail sent. | 20:23 |
| openstackgerrit | Bryan D. Payne proposed a change to openstack-infra/config: Add openstack-security to accessbot and eavesdrop https://review.openstack.org/103650 | 20:24 |
| bdpayne | jeblair https://review.openstack.org/103650 | 20:25 |
| *** homeless has quit IRC | 20:25 | |
| *** fifieldt has quit IRC | 20:25 | |
| openstackgerrit | Andreas Jaeger proposed a change to openstack-infra/config: Cleanup manual jobs https://review.openstack.org/103646 | 20:26 |
| *** e0ne has joined #openstack-infra | 20:26 | |
| *** homeless has joined #openstack-infra | 20:26 | |
| *** fifieldt has joined #openstack-infra | 20:26 | |
| *** sarob has quit IRC | 20:26 | |
| fungi | okay, i'm ready to tag a git-review release and send a release announcement (we don't have any outstanding changes that address bugs in the utility itself now, just its tests). however, since this is a major shift in packaging (switch to pbr/module, adds external dependencies, functional testing framework) and adds major features (particularly http support), should it be 2.0? commence to | 20:27 |
| fungi | bikeshedding... | 20:27 |
| clarkb | fungi: it maintains backward compat so I think next 1.x is fine | 20:28 |
| *** Longgeek has quit IRC | 20:28 | |
| *** e0ne has quit IRC | 20:28 | |
| *** yfried_ has quit IRC | 20:29 | |
| fungi | do we at least need a news file entry mentioning the changes to packaging, or just assume distro packagers will figure it out on their own? | 20:29 |
| *** yfried_ has joined #openstack-infra | 20:29 | |
| clarkb | sdague: http://logs.openstack.org/88/99688/7/check/check-tempest-dsvm-neutron-heat-slow/f3c3cec/logs/devstacklog.txt.gz?level=INFO is another one that seems to cause trouble | 20:29 |
| clarkb | sdague: I notice that the end of the log isn't timestamped | 20:29 |
| clarkb | sdague: but other than that nothing too crazy jumps out | 20:29 |
| mordred | fungi: me too (re 103355) | 20:30 |
| clarkb | mordred: are you watching algeria take it to germany? | 20:30 |
| fungi | mordred: i only gave it a +1 because i was having a bit of trouble matching up the jjb config changes to the changes in the slave scripts, but in general i'm still very excited | 20:30 |
| mordred | clarkb: EFAIL - no! I was on the phone. now it's on | 20:31 |
| sdague | clarkb: does thas really impact things? | 20:31 |
| clarkb | sdague: no idea | 20:31 |
| clarkb | sdague: but it took 14 minutes to index that one file | 20:31 |
| mordred | fungi: I _think_ I did it right :) | 20:31 |
| sdague | clarkb: I wonder if it's about the new curl in there - http://logs.openstack.org/88/99688/7/check/check-tempest-dsvm-neutron-heat-slow/f3c3cec/logs/devstacklog.txt.gz?level=INFO#_2014-06-30_19_56_51_005 | 20:32 |
| sdague | also, that's probably badness | 20:33 |
| clarkb | maybe? | 20:33 |
| fungi | hrm... at least in debian, git-review is collaboratively maintained by the openstack packaging team, so at least on debuntu this will probably just work itself out since they're already used to pbr at this point | 20:33 |
| openstackgerrit | Ilya Shakhat proposed a change to openstack-infra/config: Enable py33 gating for Stackalytics https://review.openstack.org/103654 | 20:34 |
| fungi | i'll just make sure to mention the new packaging highlights in the release announcement to cover our bases there | 20:34 |
| *** johnthetubaguy is now known as zz_johnthetubagu | 20:34 | |
| *** Sukhdev has quit IRC | 20:35 | |
| *** crc32 has quit IRC | 20:35 | |
| *** e0ne has joined #openstack-infra | 20:36 | |
| sdague | clarkb: I wonder if that came in in a way that nodepool doesn't understand it | 20:36 |
| sdague | I might not have been diligent enough on that review | 20:36 |
| *** markmcclain has joined #openstack-infra | 20:36 | |
| clarkb | sdague: maybe? we do depend on devstack to tell us what to cache now | 20:37 |
| clarkb | sdague: via that script in devstack | 20:37 |
| *** SumitNaiksatam has quit IRC | 20:38 | |
| sdague | yeh | 20:40 |
| sdague | I integrated this wrong | 20:40 |
| *** msabramo has quit IRC | 20:42 | |
| *** e0ne has quit IRC | 20:42 | |
| *** crc32 has joined #openstack-infra | 20:43 | |
| *** amotoki has quit IRC | 20:47 | |
| *** amotoki has joined #openstack-infra | 20:47 | |
| *** amotoki has quit IRC | 20:47 | |
| clarkb | hmm oslo.rootwrap doesn't seemto have docs | 20:48 |
| *** amotoki has joined #openstack-infra | 20:48 | |
| *** atiwari has quit IRC | 20:49 | |
| *** denis_makogon has quit IRC | 20:49 | |
| *** dprince has quit IRC | 20:51 | |
| *** Sukhdev has joined #openstack-infra | 20:52 | |
| *** atiwari has joined #openstack-infra | 20:52 | |
| clarkb | oslosphinx too | 20:53 |
| openstackgerrit | A change was merged to openstack-infra/config: Make ironic-undercloud voting and replace seed https://review.openstack.org/103434 | 20:56 |
| *** salv-orlando has joined #openstack-infra | 20:58 | |
| openstackgerrit | A change was merged to openstack-infra/config: Moved Tempest regex outside pre_test_hook https://review.openstack.org/103233 | 20:59 |
| *** weshay has quit IRC | 20:59 | |
| *** jdahlin has joined #openstack-infra | 21:00 | |
| *** radez is now known as radez_g0n3 | 21:01 | |
| * Ajaeger would appreciate a quick review of gating for the new security-doc repository - https://review.openstack.org/103578 | 21:01 | |
| Ajaeger | jeblair: thanks for your review! | 21:01 |
| *** yamahata has joined #openstack-infra | 21:02 | |
| openstackgerrit | Clark Boylan proposed a change to openstack-infra/config: Move oslo default test node to trusty. https://review.openstack.org/103660 | 21:04 |
| clarkb | dhellmann: ^ | 21:04 |
| clarkb | I tested py27, pep8 and doc jobs | 21:04 |
| *** SumitNaiksatam has joined #openstack-infra | 21:06 | |
| Ajaeger | fungi, thanks a lot! | 21:06 |
| clarkb | dhellmann: and since the other jobs run on dsvm they are already on trusty | 21:07 |
| anteaya | jaypipes: thanks, will read in a few | 21:07 |
| anteaya | house got hit (I think) | 21:07 |
| *** davidlenwell_ is now known as davidlenwell | 21:07 | |
| anteaya | I lost my phone but strangely have internet | 21:07 |
| anteaya | still assessing | 21:08 |
| openstackgerrit | A change was merged to openstack-infra/config: Add gating for security-doc repo https://review.openstack.org/103578 | 21:10 |
| *** msabramo has joined #openstack-infra | 21:12 | |
| clarkb | mordred: in 103363 you still need to select a mirror | 21:12 |
| clarkb | mordred: otherwise how will it know to use your new bandersnatch mirror? | 21:12 |
| clarkb | mordred: those cp's will copy empty files | 21:13 |
| clarkb | or non existant files | 21:13 |
| mordred | clarkb: why? | 21:13 |
| mordred | clarkb: did you look at the part of that patch where we install base config files that are not empty? | 21:13 |
| fungi | clarkb: finish looking at the change, those files are moved | 21:13 |
| clarkb | mordred: because the tool default is pypi.python.org | 21:13 |
| mordred | clarkb: look at the rest of the patch | 21:14 |
| fungi | clarkb: and installed permenantly into the slave | 21:14 |
| fungi | permanently | 21:14 |
| clarkb | I am looking at the whole thing | 21:14 |
| clarkb | is there a different change? | 21:14 |
| mordred | clarkb: in the config sequence | 21:15 |
| mordred | sorry - you were looking at the devstack-gate change .. I should have mentioned it needs the other change to land first | 21:15 |
| clarkb | yes please :) | 21:15 |
| fungi | heh, i missed the d-g change... which one is that? | 21:15 |
| *** sarob has joined #openstack-infra | 21:15 | |
| mordred | https://review.openstack.org/#/c/103363 | 21:16 |
| mordred | it's all topic bandersnatch | 21:16 |
| mordred | clarkb: https://review.openstack.org/#/c/103355/ this is where the files get installed | 21:16 |
| *** sabeen has quit IRC | 21:19 | |
| clarkb | arg office wifi so bad | 21:19 |
| *** markmcclain has quit IRC | 21:20 | |
| *** julim has quit IRC | 21:21 | |
| clarkb | I shouldnt need to tether in the office... | 21:21 |
| *** doddstack has joined #openstack-infra | 21:24 | |
| *** thedodd has quit IRC | 21:25 | |
| *** maxbit has joined #openstack-infra | 21:25 | |
| *** HenryG_ has joined #openstack-infra | 21:26 | |
| *** homeless_ has joined #openstack-infra | 21:26 | |
| *** pdmars has quit IRC | 21:27 | |
| *** hashar_ has joined #openstack-infra | 21:27 | |
| *** nati_uen_ has joined #openstack-infra | 21:27 | |
| openstackgerrit | Andreas Jaeger proposed a change to openstack-infra/config: Fix typo: security-guide -> security-doc https://review.openstack.org/103664 | 21:27 |
| mordred | clarkb: I tether at the office all the time | 21:27 |
| clarkb | mordred: another question for you. pip.conf and pydistutils.cfg files are moved to openstack_project. Were they used at all in jenkins? | 21:27 |
| clarkb | mordred: its crazy that that would be a thing | 21:28 |
| *** EntropyWorks_ has joined #openstack-infra | 21:28 | |
| *** radez_g0` has joined #openstack-infra | 21:28 | |
| clarkb | mordred: that seems like an excellent "this is why I work at home" response | 21:28 |
| *** rcarrill` has joined #openstack-infra | 21:28 | |
| *** yamahata__ has joined #openstack-infra | 21:29 | |
| mordred | clarkb: they were not | 21:29 |
| mordred | clarkb: which I found amusing | 21:29 |
| *** mikap has joined #openstack-infra | 21:31 | |
| *** jogo_awa1 has joined #openstack-infra | 21:31 | |
| *** Shrews_ has joined #openstack-infra | 21:32 | |
| jgriffith | jeblair: clarkb any idea why this might be happening: http://logs.openstack.org/26/103426/1/check/gate-cinder-python27/7844c61/console.html#_2014-06-29_19_21_12_875 | 21:32 |
| jgriffith | jeblair: clarkb never mind | 21:33 |
| *** andreykurilin has joined #openstack-infra | 21:33 | |
| openstackgerrit | Michael Krotscheck proposed a change to openstack-infra/infra-specs: StoryBoard Permissions Specifications https://review.openstack.org/103667 | 21:34 |
| *** sarob has quit IRC | 21:34 | |
| *** EntropyWorks has quit IRC | 21:34 | |
| *** EntropyWorks_ is now known as EntropyWorks | 21:34 | |
| *** radez_g0n3 has quit IRC | 21:34 | |
| *** rcarrillocruz has quit IRC | 21:34 | |
| stevebaker | hey, there are 22 check jobs which haven't terminated for 16 hours (21 of them are mine ;) http://status.openstack.org/zuul/ | 21:34 |
| *** HenryG has quit IRC | 21:34 | |
| jesusaurus | clarkb: did you get a chance to look at https://review.openstack.org/#/c/103258/? | 21:35 |
| clarkb | jesusaurus: is that the one about apache? | 21:35 |
| *** ubuntu has joined #openstack-infra | 21:35 | |
| clarkb | jesusaurus: I don't think that was the only problem with 0.4.0 | 21:35 |
| jesusaurus | yeah, reverting your downgrade to 0.0.4 | 21:35 |
| *** homeless has quit IRC | 21:35 | |
| *** eharney has quit IRC | 21:35 | |
| *** nati_ueno has quit IRC | 21:35 | |
| *** jgrimm has quit IRC | 21:35 | |
| *** Shrews has quit IRC | 21:35 | |
| *** jogo_away has quit IRC | 21:35 | |
| *** kevinbenton has quit IRC | 21:35 | |
| *** mika has quit IRC | 21:35 | |
| *** zaro has quit IRC | 21:35 | |
| *** maxbit_ has quit IRC | 21:35 | |
| *** hashar has quit IRC | 21:35 | |
| *** Daviey has quit IRC | 21:35 | |
| *** juice has quit IRC | 21:35 | |
| *** yamahata has quit IRC | 21:35 | |
| *** greghaynes has quit IRC | 21:35 | |
| *** Ajaeger has quit IRC | 21:36 | |
| *** ubuntu is now known as Guest37445 | 21:36 | |
| clarkb | jesusaurus: and if we are going to change we should go to whatever is currentish | 21:36 |
| *** mikap is now known as mika | 21:36 | |
| clarkb | but that may be a bigger delta | 21:36 |
| fungi | stevebaker: something happened to the network connectivity for jenkins03, so jobs which were running on it at the time got their statuses forever lost. the options there are currently either to push a new patchset to those or restart zuul | 21:36 |
| *** Ajaeger has joined #openstack-infra | 21:36 | |
| jesusaurus | clarkb: well, we at least need to specify a custom vhost template, so i think 0.0.7 (i would need to double check) is the newest we could do | 21:36 |
| *** juice has joined #openstack-infra | 21:36 | |
| *** kevinbenton has joined #openstack-infra | 21:36 | |
| stevebaker | fungi: ok, I can push a new changeset | 21:37 |
| clarkb | jesusaurus: wait did they remove that feature again? | 21:37 |
| jesusaurus | again? they removed it long ago | 21:37 |
| *** Daviey has joined #openstack-infra | 21:37 | |
| jesusaurus | or did they re-add it and i didnt notice? | 21:37 |
| clarkb | jesusaurus: I thin kthey readded it because I complained | 21:37 |
| clarkb | and everyone else | 21:37 |
| clarkb | nibalizer: ^ | 21:37 |
| Ajaeger | fungi, jeblair I overlooked one thing in my last patch, could you review https://review.openstack.org/103664, please? | 21:37 |
| nibalizer | you can custom_fragment now | 21:38 |
| jgriffith | mordred: so none of the oslo requirements in the globa-requirements file are pinned with upper bounds? | 21:38 |
| jgriffith | for stable/icehouse | 21:39 |
| sdague | jgriffith: right, that's general policy though | 21:39 |
| openstackgerrit | pritesh proposed a change to openstack-dev/hacking: Add hacking check for editor configuration in source files https://review.openstack.org/101969 | 21:39 |
| jgriffith | sdague: but then shit breaks | 21:39 |
| jgriffith | sdague: ie I get that on master | 21:39 |
| jgriffith | sdague: but for stable? | 21:39 |
| fungi | jgriffith: nah. for those, you know where to send hate mail (burning poop, cetera) if they break you ;) | 21:39 |
| *** dkranz has quit IRC | 21:39 | |
| jgriffith | fungi: well I get plenty from Mark :) | 21:39 |
| sdague | jgriffith: well, markmc is actually one of the strongest advocates for not pinning | 21:40 |
| *** whoops has joined #openstack-infra | 21:40 | |
| jgriffith | so I'm in a perdicament with the oslo.messaging update | 21:40 |
| clarkb | nibalizer: oh so you can't do a realy template? | 21:40 |
| jgriffith | sdague: what I mean is... | 21:40 |
| clarkb | nibalizer: jesusaurus: I cannot grok how that is an improvement | 21:40 |
| jgriffith | if you look at the ML thread between he and I just now | 21:40 |
| jesusaurus | nibalizer: oh cool, and custom_fragments would allow us to do mod_rewrite proxy stuff? how big of a change would that be? | 21:40 |
| sdague | jgriffith: yeh, I was just reading it | 21:40 |
| jgriffith | oslo.messaging update changes config | 21:40 |
| clarkb | nibalizer: jesusaurus: that is like apple taking away features from its users | 21:40 |
| clarkb | then saying yo ushould be happy for it | 21:41 |
| jgriffith | sdague: ok... I'll refrain from paraphrasing | 21:41 |
| jesusaurus | clarkb: but you WANT a macbook wheel | 21:41 |
| sdague | so honestly, it's not helpful, but the libraries having config vars is where this really breaks down | 21:42 |
| *** sarob has joined #openstack-infra | 21:42 | |
| jgriffith | sdague: yeah... as we now see | 21:42 |
| sdague | honestly, I'm ok with proposing that all oslo libraries are pinned | 21:42 |
| *** eharney has joined #openstack-infra | 21:42 | |
| *** msabramo has quit IRC | 21:42 | |
| sdague | as the oslo libraries are apparently only going to release with the release | 21:42 |
| *** mtreinish has joined #openstack-infra | 21:42 | |
| *** blamar has quit IRC | 21:43 | |
| jgriffith | sdague: I'm at least going to propose pinning stable/icehouse oslo.messaging at the moment | 21:43 |
| sdague | so they should be set to >=1.3.0,<1.4.0 on stable/icehouse | 21:43 |
| *** arnaud__ has joined #openstack-infra | 21:43 | |
| sdague | jgriffith: yeh, it would probably be good to force policy back there on the list as well | 21:43 |
| jgriffith | agreed, trying to figure out where "everything" else should be | 21:43 |
| jgriffith | sdague: ok.. thanks | 21:43 |
| jgriffith | I'll send a thread ot ML and log a defect | 21:44 |
| *** greghaynes has joined #openstack-infra | 21:44 | |
| *** arnaud has quit IRC | 21:44 | |
| *** arnaud has joined #openstack-infra | 21:45 | |
| fungi | jgriffith: were you able to figure out why it was impacting py27 tests but not py26? | 21:45 |
| *** annegent_ has quit IRC | 21:45 | |
| cody-somerville | What causes pypi mirror to create Ubuntu-12.04 vs. Ubuntu-13.10 directory? | 21:46 |
| fungi | that definitely felt like a spooky-action-at-a-distance sort of thing | 21:46 |
| *** annegent_ has joined #openstack-infra | 21:46 | |
| clarkb | cody-somerville: its part of the pypi mirror tool | 21:46 |
| fungi | cody-somerville: the wheel-building depends on the platform on which it's run | 21:46 |
| clarkb | cody-somerville: it builds platform specific wheels | 21:46 |
| jgriffith | fungi: not quite yet | 21:47 |
| jgriffith | had to step away for a bit | 21:47 |
| cody-somerville | clarkb: fungi: Is it possible to get both created? | 21:47 |
| fungi | cody-somerville: yes, by running on both and rsync'ing them to a common destination | 21:47 |
| *** james_li has joined #openstack-infra | 21:47 | |
| jesusaurus | clarkb: fungi: where is that being run? on the pypi-mirror host? or can the wheels be built elsewhere then published? | 21:48 |
| * jesusaurus needs to better grok wheels | 21:48 | |
| clarkb | jesusaurus: it is run where you run the pypi mirror tool | 21:48 |
| clarkb | we use a jenkins slave | 21:48 |
| fungi | jesusaurus: it's happening on our mirror26/27/33 jenjins slaves right now | 21:48 |
| clarkb | these slaves rsync to the common mirror location | 21:48 |
| clarkb | also please stop using 13.10 that is all | 21:48 |
| fungi | clarkb: sts is the new lts! | 21:49 |
| jesusaurus | its not eol for... a couple weeks... | 21:49 |
| nibalizer | clarkb: so you can custom_fragment your way to victory | 21:49 |
| clarkb | jesusaurus: right so spend your time now moving to trusty instead | 21:49 |
| clarkb | nibalizer: I shouldn't need to | 21:49 |
| jesusaurus | clarkb: ++ | 21:49 |
| *** dangers is now known as dangers_away | 21:50 | |
| clarkb | mordred: please see comment on 103355 | 21:51 |
| *** annegent_ has quit IRC | 21:51 | |
| fungi | puppet templates are the puppet answer to customizing a file. interspersed/interleaved file fragments seem very un-puppetly | 21:51 |
| nibalizer | clarkb: so you can go custom_fragment => template(thingy.vhost.erb), | 21:52 |
| nibalizer | you'll have to set listen and snip the bottom and top lines of the vhost but other than that, should work pretty well | 21:52 |
| jesusaurus | fungi: i disagree, i think that things like the concat module are a very puppetty answer to config files that arent very config-management friendly | 21:53 |
| *** whoops has quit IRC | 21:53 | |
| *** whoops has joined #openstack-infra | 21:53 | |
| *** arnaud has quit IRC | 21:54 | |
| nibalizer | fungi: my point is we can replace vhost_template => template('thingy.vhost.erb') with custom_fragment => template('thingy.vhost.erb') | 21:55 |
| *** arnaud has joined #openstack-infra | 21:55 | |
| clarkb | nibalizer: and rewrite all of our vhosts | 21:55 |
| nibalizer | so we can use the modern apache module much as the old one was | 21:55 |
| nibalizer | rewrite is a stronger word than i would use | 21:55 |
| clarkb | in a way that shouldn't be necessary | 21:55 |
| nibalizer | you just snip off the listen component and thats it | 21:55 |
| clarkb | because the first release supported the only thing you really need... | 21:55 |
| *** hashar_ is now known as hashar | 21:55 | |
| *** andreaf_ has quit IRC | 21:56 | |
| fungi | jesusaurus: well, we at least agree that apache's httpd.conf isn't very config-management friendly ;) | 21:56 |
| *** r1chardj0n3s_afk is now known as r1chardj0n3s | 21:56 | |
| *** jamielennox is now known as jamielennox|away | 21:56 | |
| nibalizer | the older one was loosely modeled, the new one is tightly modeled | 21:56 |
| nibalizer | but you can still pass arbitrary stuff into the vhost and apache module will just get-it-done | 21:56 |
| nibalizer | would it be helpful if i made a work-in-progress review showing what the implementaion of the new apache module would look like? | 21:58 |
| clarkb | nibalizer: I think we know what it would look like | 21:58 |
| *** chuckC has quit IRC | 21:58 | |
| clarkb | nibalizer: the frustration is that it is necessary and that there is no good way to make the change piecemeal | 21:58 |
| nibalizer | ya | 21:58 |
| nibalizer | picemeal would be better | 21:58 |
| clarkb | so its all or nothing | 21:59 |
| *** dims has quit IRC | 22:00 | |
| *** pballand has quit IRC | 22:02 | |
| *** marcoemorais has quit IRC | 22:03 | |
| *** marcoemorais has joined #openstack-infra | 22:03 | |
| *** amotoki has quit IRC | 22:03 | |
| *** markwash has quit IRC | 22:04 | |
| *** andreykurilin has quit IRC | 22:05 | |
| *** lcheng has quit IRC | 22:05 | |
| *** jhesketh has quit IRC | 22:06 | |
| *** doug-fish has left #openstack-infra | 22:06 | |
| fungi | mordred: you may want to follow up to http://lists.openstack.org/pipermail/openstack-dev/2014-June/039023.html with any relevant info on how your mirror changes will benefit the prerelease/wheel testing situation, lest the void be filled with misinfo | 22:07 |
| mattoliverau | Morning | 22:08 |
| clarkb | mattoliverau: good morning | 22:09 |
| clarkb | hrm no jhesketh yet | 22:09 |
| * clarkb waits patiently | 22:09 | |
| *** Ajaeger has quit IRC | 22:09 | |
| *** lcheng has joined #openstack-infra | 22:09 | |
| *** sarob has quit IRC | 22:10 | |
| mattoliverau | Lol, not yet. | 22:10 |
| *** sarob has joined #openstack-infra | 22:10 | |
| *** pballand has joined #openstack-infra | 22:11 | |
| *** markwash has joined #openstack-infra | 22:11 | |
| *** melwitt has quit IRC | 22:12 | |
| *** sarob has quit IRC | 22:15 | |
| *** HenryG_ has quit IRC | 22:16 | |
| clarkb | dhellmann: any chance I can get you to weigh in on https://review.openstack.org/#/c/103660/ | 22:16 |
| clarkb | dhellmann: oslo.config oslo.db oslo.i18n oslo-incubator oslo.messaging oslo.rootwrap oslo-specs oslosphinx oslotest oslo.version oslo.vmware those are the projects I tested | 22:16 |
| *** markwash has quit IRC | 22:17 | |
| *** hashar has quit IRC | 22:17 | |
| *** markwash has joined #openstack-infra | 22:18 | |
| ianw | sdague: https://review.openstack.org/#/c/100131/ (Make one copy of grenade log files) still seems to be happening | 22:18 |
| *** dkliban is now known as dkliban_afk | 22:19 | |
| *** HenryG has joined #openstack-infra | 22:21 | |
| *** HenryG has quit IRC | 22:21 | |
| *** camunoz_gone is now known as camunoz | 22:21 | |
| openstackgerrit | A change was merged to openstack-infra/devstack-gate: Archive Ironic VM nodes console logs https://review.openstack.org/92216 | 22:21 |
| *** gondoi is now known as zz_gondoi | 22:21 | |
| *** HenryG has joined #openstack-infra | 22:22 | |
| anteaya | lost phone line at the house, next door lost phone line and hot tub | 22:23 |
| anteaya | 4.75 inches of water in 2 hours | 22:23 |
| *** hashar has joined #openstack-infra | 22:23 | |
| *** hashar has quit IRC | 22:23 | |
| clarkb | anteaya: wow | 22:23 |
| anteaya | don't understand how we lost phone line and still have internet | 22:23 |
| anteaya | clarkb: yeah, I'm amazed | 22:24 |
| anteaya | two storms | 22:24 |
| fungi | the hot tub floated away? | 22:24 |
| anteaya | we never get weather like this | 22:24 |
| anteaya | fungi: ground fault | 22:24 |
| anteaya | it got hit | 22:24 |
| anteaya | or something did | 22:24 |
| sdague | ianw: +2 | 22:24 |
| openstackgerrit | A change was merged to openstack-infra/config: Fix typo: security-guide -> security-doc https://review.openstack.org/103664 | 22:24 |
| nibalizer | anteaya: where are you based? | 22:24 |
| anteaya | haliburton | 22:24 |
| whoops | I'm guessing you have cable internet.. the line is probably buried | 22:24 |
| *** msabramo has joined #openstack-infra | 22:24 | |
| *** jaypipes has quit IRC | 22:24 | |
| anteaya | https://en.wikipedia.org/wiki/Dysart_et_al,_Ontario | 22:25 |
| anteaya | whoops: no dsl | 22:25 |
| anteaya | and isp checked the line, only 10 decibels of random noise | 22:25 |
| whoops | really? then it's really odd you have internet but no phone | 22:25 |
| anteaya | they do something at 30 decibels | 22:25 |
| anteaya | whoops: yeah, same at the other house too | 22:26 |
| anteaya | really weird | 22:26 |
| bknudson | it's a canada day miracle | 22:26 |
| anteaya | bknudson: praise be | 22:26 |
| *** rdopieralski has quit IRC | 22:27 | |
| fungi | all glory to the dominion? | 22:27 |
| openstackgerrit | Monty Taylor proposed a change to openstack-infra/config: Stop running pypi-mirror for old mirror creation https://review.openstack.org/103360 | 22:27 |
| openstackgerrit | Monty Taylor proposed a change to openstack-infra/config: Make select-mirror use the bandersnatch mirror https://review.openstack.org/103354 | 22:27 |
| openstackgerrit | Monty Taylor proposed a change to openstack-infra/config: Delete select-mirror and all the args https://review.openstack.org/103355 | 22:27 |
| *** virmitio has left #openstack-infra | 22:27 | |
| anteaya | fungi: welll, not sure I'd go that far | 22:27 |
| clarkb | fungi: hypnotoad | 22:27 |
| mordred | fungi, clarkb, jeblair: ^^ I reordered these to allow us to more safely land the next in the series without removing the infrastructure needed ot rollback | 22:27 |
| anteaya | but glad I can still work, while the three stoges phone company takes the rest of the summer to fix service | 22:28 |
| fungi | later this week my country will celebrate its independence by blowing up very small pieces of itself. you know, for fun | 22:29 |
| *** marcoemorais has quit IRC | 22:30 | |
| anteaya | yeah | 22:30 |
| anteaya | so there's that to look forward to | 22:30 |
| fungi | <simpsons>celebrate the independence of your nation by blowing up a small part of it</simpsons> | 22:30 |
| *** marcoemorais has joined #openstack-infra | 22:30 | |
| anteaya | I have had fireworks for three nights straight | 22:30 |
| anteaya | based on the number of cottagers hanging out | 22:30 |
| anteaya | I'd put money on two more nights of it | 22:30 |
| *** marcoemorais has quit IRC | 22:31 | |
| anteaya | fungi: ah, see I don't watch the simpsons | 22:31 |
| anteaya | that is my nephew's | 22:31 |
| *** marcoemorais has joined #openstack-infra | 22:31 | |
| anteaya | I can't infringe | 22:31 |
| *** blamar has joined #openstack-infra | 22:32 | |
| fungi | yeah, that was a reference to an episode from ~18 years ago | 22:32 |
| *** james_li has quit IRC | 22:33 | |
| *** marcoemorais has quit IRC | 22:33 | |
| *** sarob has joined #openstack-infra | 22:33 | |
| *** flaper87 is now known as flaper87|afk | 22:34 | |
| jeblair | fungi: i don't remember that quote, which suggests it's probably time for me to start over again from the beginning ;) | 22:34 |
| *** marcoemorais has joined #openstack-infra | 22:34 | |
| anteaya | fungi: ah, I really wouldn't have gotten it then | 22:34 |
| anteaya | jeblair: ha ha ha | 22:34 |
| *** pballand has quit IRC | 22:34 | |
| anteaya | does a simpsons fan need an excuse to watch simpsons? | 22:34 |
| clarkb | jeblair: for https://review.openstack.org/#/c/103019/3 that will stop running those jobs with the jenkins provided maven and use system provided maven. Did you take that into account with your +2? | 22:35 |
| clarkb | jeblair: mostly wondering if that was discussed somewhere and I should just not worry about it | 22:35 |
| * anteaya considers food | 22:35 | |
| openstackgerrit | A change was merged to openstack-infra/config: Remove envinject from gerrit-plugin-jobs https://review.openstack.org/102087 | 22:35 |
| clarkb | jeblair: if we don't think that is a problem I will go ahead and approve | 22:35 |
| openstackgerrit | A change was merged to openstack-infra/config: Remove envinject from gerrit https://review.openstack.org/102088 | 22:36 |
| fungi | jeblair: season 7, episode 25, "summer of 4 ft. 2" | 22:36 |
| openstackgerrit | A change was merged to openstack-infra/config: Remove envinject from javascript https://review.openstack.org/102089 | 22:36 |
| jeblair | clarkb: i'm not positive, but it's something we should work toward anyway | 22:36 |
| clarkb | jeblair: ok so better to discover a problem with it now than later wfm | 22:36 |
| clarkb | (and ya with future gearman world that makes sense | 22:36 |
| jeblair | clarkb: yeah, and i think our tolerance for discovering problems in that space is high right now | 22:37 |
| *** msabramo has quit IRC | 22:37 | |
| fungi | it's not feature freeze rush *yet* | 22:37 |
| clarkb | approved | 22:37 |
| *** pballand has joined #openstack-infra | 22:38 | |
| *** pballand has quit IRC | 22:38 | |
| openstackgerrit | A change was merged to openstack-infra/config: Remove envinject from groups https://review.openstack.org/102090 | 22:39 |
| openstackgerrit | Clark Boylan proposed a change to openstack-infra/config: Move trove default test node to trusty. https://review.openstack.org/103678 | 22:40 |
| openstackgerrit | Clark Boylan proposed a change to openstack-infra/config: Move troveclient default test node to trusty. https://review.openstack.org/103679 | 22:40 |
| clarkb | MOAR trusty | 22:40 |
| clarkb | mordred: any word on the volume stuff? | 22:40 |
| *** rlandy has quit IRC | 22:41 | |
| *** dimtruck is now known as zz_dimtruck | 22:42 | |
| *** zzelle_ has quit IRC | 22:42 | |
| SlickNik | clarkb: nice! | 22:42 |
| *** doddstack has quit IRC | 22:43 | |
| mordred | clarkb: yes. one sec | 22:44 |
| openstackgerrit | A change was merged to openstack-infra/config: Remove envinject usage from maven-plugin-jobs.yaml https://review.openstack.org/103019 | 22:45 |
| jgriffith | sdague: fungi https://review.openstack.org/#/c/103674/ | 22:45 |
| jgriffith | fungi: it would seem that that alpha build isn't py26 compat so isn't used. Haven't figured out quite how it figures that out | 22:46 |
| *** msabramo has joined #openstack-infra | 22:46 | |
| clarkb | mordred: I don't think latest patchset of https://review.openstack.org/#/c/103355/ addressed my comment | 22:46 |
| fungi | jgriffith: yep, saw the ml thread. +1 but i'm not stable branch core | 22:46 |
| jgriffith | fungi: but that *appears* to be the deal | 22:46 |
| jgriffith | fungi: ml? | 22:46 |
| * jgriffith looks | 22:46 | |
| jgriffith | oh... derp | 22:47 |
| jgriffith | yeah | 22:47 |
| jgriffith | :) | 22:47 |
| fungi | jgriffith: oh, well i saw markmc's comment about the prerelease version spec in the stable reqs files being at fault for pulling in the new alpha weel | 22:47 |
| fungi | wheel | 22:47 |
| fungi | the py26 bit is still news to me, and will keep that in mind if we see similar in the future, so thanks! | 22:47 |
| *** marcoemorais has quit IRC | 22:48 | |
| jgriffith | fungi: certainly | 22:48 |
| *** marcoemorais has joined #openstack-infra | 22:48 | |
| fungi | jgriffith: i'm guessing though that it's a python 2.7 version expressed in the wheel | 22:48 |
| *** r1chardj0n3s is now known as r1chard_afk | 22:48 | |
| jgriffith | fungi: makes sense... I'll buy it and move on :) | 22:48 |
| clarkb | jeblair: any particular reason you didn't approve 103421? | 22:48 |
| mordred | clarkb: ++ | 22:49 |
| fungi | jgriffith: hrm, though the whl file just claims to be py2 not py27... http://pypi.openstack.org/openstack/oslo.messaging/oslo.messaging-1.4.0.0a2-py2-none-any.whl | 22:49 |
| fungi | so no idea really | 22:49 |
| openstackgerrit | A change was merged to openstack-infra/config: Remove Jenkins EnvInject plugin from jenkins_dev https://review.openstack.org/101095 | 22:49 |
| *** devkulkarni has quit IRC | 22:50 | |
| dstufft | wat | 22:50 |
| jeblair | clarkb: didn't want to deal with the consequences | 22:50 |
| clarkb | mordred: for 103421, do you need to include -cpp as well? | 22:50 |
| clarkb | jeblair: :) ok | 22:50 |
| clarkb | mordred: I think that is what the c++ resource was meant to be | 22:51 |
| clarkb | as I don't see a c++ compiler | 22:51 |
| clarkb | only cpp and g++ | 22:51 |
| *** Sukhdev has quit IRC | 22:51 | |
| clarkb | I guess that could be platform dependent | 22:51 |
| clarkb | I will leave a comment | 22:52 |
| *** whoops has quit IRC | 22:52 | |
| mordred | clarkb: ccache doesn't need to be set for cpp | 22:53 |
| mordred | clarkb: ccache works on the results of the output of the preprocessor stage | 22:53 |
| *** annegent_ has joined #openstack-infra | 22:53 | |
| clarkb | oh it doesn't cache any of the preprocessor data? | 22:54 |
| *** sarob has quit IRC | 22:54 | |
| *** homeless_ has quit IRC | 22:54 | |
| *** gargola_ has joined #openstack-infra | 22:54 | |
| *** W00die has joined #openstack-infra | 22:54 | |
| *** hemna__ has joined #openstack-infra | 22:54 | |
| *** dangers_` has joined #openstack-infra | 22:54 | |
| clarkb | well then I shall approve instead | 22:54 |
| mordred | it caches the result, aiui - but in either case, it does not hook into cpp | 22:54 |
| clarkb | roger | 22:54 |
| dstufft | fungi: basically the whl file stuff is supposed to define the outcome of setup.py, so even if something doesn't work on python 2.4 (for instance), if setup.py would have the same outcome on python2.4 and 2.7 then it should be py2 | 22:54 |
| *** gargola has quit IRC | 22:54 | |
| *** hemna_ has quit IRC | 22:55 | |
| *** Woodie has quit IRC | 22:55 | |
| *** dangers_away has quit IRC | 22:55 | |
| *** homeless has joined #openstack-infra | 22:55 | |
| *** msabramo has quit IRC | 22:55 | |
| fungi | dstufft: yeah, reasoning about why a centos6 machine didn't pull in a wheel of a prerelease and fell back to the release tarball, while an ubuntu 14.04 machine grabbed the wheel | 22:55 |
| * clarkb gets brave and approves things that jeblair didn't want to deal with consequences on | 22:55 | |
| fungi | dstufft: mostly grasping at straws | 22:56 |
| dstufft | also wheels are generally implemented as"and later than" | 22:56 |
| dstufft | so a py26 wheel will be used on py27 | 22:56 |
| dstufft | If I recall | 22:56 |
| * fungi worries that a py27 wheel would be used on py31? | 22:56 | |
| dstufft | nah | 22:57 |
| dstufft | it's smarter than that | 22:57 |
| fungi | 2.and-later-than vs 3.and-later-than i guess | 22:57 |
| dstufft | pretty sure that's special cased for pure python wheels too | 22:57 |
| fungi | got it | 22:57 |
| dstufft | the idea being that the likelyhood that somehting for 3.3 also works on 3.4 if it's pure python is pretty high | 22:57 |
| *** mmaglana has quit IRC | 22:58 | |
| *** mmaglana has joined #openstack-infra | 22:58 | |
| mordred | fungi, clarkb: ok. I'm going to land the select_mirror patch | 22:59 |
| *** atiwari has quit IRC | 23:00 | |
| mordred | unless anyone has any last minute concerns... | 23:00 |
| clarkb | mordred: nope, just remember you need to rebuild images to see the change | 23:00 |
| mordred | oh right. so landing teh patch is safeish for now | 23:00 |
| clarkb | yes | 23:01 |
| fungi | mordred: just please be around at ~14:00 utc tomorrow when the image rebuilds happen, if you're not planning to trigger fresh builds tonight and watch jobs run on the resulting nodes ;) | 23:02 |
| clarkb | also that ^ | 23:02 |
| * fungi does not relish being on revert duty when it can be avoided | 23:02 | |
| mordred | fungi: I'm going to trigger images | 23:03 |
| * fungi is entirely in favor of this plan! | 23:03 | |
| clarkb | it doesn't look like dhellmann is/was around today | 23:04 |
| clarkb | how do people feel about pushing through the oslo on trusty change anyways? | 23:04 |
| clarkb | my biggest concern is that they are running some test that I didn't expect because they do that | 23:04 |
| clarkb | and dhellmann might be able to point that out if it exists | 23:04 |
| *** markwash has quit IRC | 23:04 | |
| clarkb | oh ENOBATTERY /me moves | 23:04 |
| *** jhesketh_ has joined #openstack-infra | 23:04 | |
| fungi | clarkb: have we yet found anything besides the reqs integration job which didn't actually work fine on trusty? | 23:05 |
| fungi | oh, and there was a solum job... something not quite right with rabbitmq apparently? | 23:05 |
| openstackgerrit | A change was merged to openstack-infra/config: Add symlinks for plat-specific compiler names https://review.openstack.org/103421 | 23:05 |
| *** jhesketh_ is now known as jhesketh | 23:06 | |
| jhesketh | Morning | 23:06 |
| *** weshay has joined #openstack-infra | 23:06 | |
| *** msabramo has joined #openstack-infra | 23:08 | |
| clarkb | fungi: thats basically it | 23:09 |
| openstackgerrit | A change was merged to openstack-infra/config: Make select-mirror use the bandersnatch mirror https://review.openstack.org/103354 | 23:09 |
| clarkb | fungi: horizon but I reproduced on precise | 23:09 |
| openstackgerrit | A change was merged to openstack-infra/config: IRC setup for security-doc and training-guides https://review.openstack.org/103597 | 23:09 |
| fungi | oh, right the httplib2 thing | 23:09 |
| fungi | yeah, seems a safe enough bet to me | 23:09 |
| *** sarob has joined #openstack-infra | 23:10 | |
| *** r1chard_afk is now known as r1chardj0n3s | 23:10 | |
| anteaya | morning jhesketh | 23:10 |
| anteaya | jhesketh: can you use you powers of +A and merge this for me, all green and already has a +2 | 23:10 |
| anteaya | https://review.openstack.org/#/c/102960/ | 23:10 |
| openstackgerrit | A change was merged to openstack-infra/config: Remove system cron activities on slaves https://review.openstack.org/103430 | 23:11 |
| *** cindyo has left #openstack-infra | 23:12 | |
| *** Nikolay_St has quit IRC | 23:12 | |
| anteaya | jhesketh: oh and this too: https://review.openstack.org/#/c/102538/ | 23:15 |
| *** devkulkarni has joined #openstack-infra | 23:15 | |
| *** sabeen1 has joined #openstack-infra | 23:16 | |
| *** annegent_ has quit IRC | 23:17 | |
| *** chuckC has joined #openstack-infra | 23:17 | |
| *** openstackgerrit has quit IRC | 23:19 | |
| *** openstackgerrit has joined #openstack-infra | 23:20 | |
| *** gargola_ has quit IRC | 23:22 | |
| *** yamahata__ has quit IRC | 23:23 | |
| *** gargola has joined #openstack-infra | 23:23 | |
| *** Ryan_Lane has quit IRC | 23:24 | |
| *** arnaud has quit IRC | 23:25 | |
| *** arnaud has joined #openstack-infra | 23:26 | |
| *** zhiyan_ is now known as zhiyan | 23:26 | |
| *** markmcclain has joined #openstack-infra | 23:26 | |
| *** dims_ has joined #openstack-infra | 23:27 | |
| *** zhiyan is now known as zhiyan_ | 23:27 | |
| clarkb | jhesketh: does 102893 conflict with your stack that I just approved? | 23:27 |
| *** jamielennox|away is now known as jamielennox | 23:28 | |
| *** zhiyan_ is now known as zhiyan | 23:29 | |
| anteaya | clarkb: yay, thank you | 23:30 |
| *** markmcclain has quit IRC | 23:30 | |
| *** zhiyan is now known as zhiyan_ | 23:30 | |
| *** Nikolay_St has joined #openstack-infra | 23:30 | |
| *** oomichi has joined #openstack-infra | 23:31 | |
| *** lcheng has quit IRC | 23:31 | |
| openstackgerrit | Clark Boylan proposed a change to openstack-infra/config: Add python34 build and job template https://review.openstack.org/102847 | 23:32 |
| clarkb | zul: ^ quick update to your change so that we can keep that ball rolling. thank you for starting it | 23:32 |
| *** dims_ has quit IRC | 23:32 | |
| openstackgerrit | A change was merged to openstack-infra/config: Proper passing of SUDO flag for neutron functional tests https://review.openstack.org/103196 | 23:33 |
| clarkb | fungi: I think 102822 will need a rebase now? | 23:34 |
| clarkb | fungi: is that somethign your script tools will have made easy? | 23:34 |
| clarkb | I suppose sed may just work (tm) | 23:34 |
| *** markmcclain has joined #openstack-infra | 23:34 | |
| jhesketh | anteaya: 102893 doesn't conflict, it's needed :-) | 23:35 |
| openstackgerrit | A change was merged to openstack-infra/config: Adds test artifact retention requirement https://review.openstack.org/102960 | 23:35 |
| clarkb | jhesketh: I thought you had a change that did that though | 23:36 |
| *** david-lyle has quit IRC | 23:36 | |
| jhesketh | clarkb: in a different module | 23:36 |
| clarkb | oh that was maven-jobs | 23:36 |
| jhesketh | I only skipped that module because Khai had done it for me | 23:37 |
| clarkb | gotcha | 23:37 |
| clarkb | approved | 23:37 |
| clarkb | with that change I think we have removed all the envinject? | 23:37 |
| *** dims_ has joined #openstack-infra | 23:37 | |
| *** Nikolay_St has quit IRC | 23:38 | |
| *** nelsnelson has quit IRC | 23:38 | |
| jhesketh | ta | 23:38 |
| openstackgerrit | A change was merged to openstack-infra/config: remove envinject usage from jenkins-plugin-jobs.yaml https://review.openstack.org/102893 | 23:40 |
| *** weshay has quit IRC | 23:41 | |
| *** kgriffs|afk has quit IRC | 23:41 | |
| *** kgriffs|afk has joined #openstack-infra | 23:41 | |
| *** kgriffs|afk is now known as kgriffs | 23:42 | |
| clarkb | jhesketh: one step closer! | 23:42 |
| openstackgerrit | A change was merged to openstack-infra/config: Clarifies that CI accounts must request accounts from infra https://review.openstack.org/102538 | 23:42 |
| clarkb | jhesketh: maybe we should take a stab at using krtaylor's hack for uploading console.html files? | 23:43 |
| jhesketh | clarkb: yeah I like that... I've asked if he can perhaps incorporate his changes and use our other files | 23:43 |
| clarkb | jhesketh: we can just have your python script fork a child to do that right? | 23:44 |
| openstackgerrit | A change was merged to openstack-infra/config: Add missing export -f in rally-cli job https://review.openstack.org/102834 | 23:44 |
| clarkb | or is it shell? | 23:44 |
| clarkb | in either case the method should work regardless of language | 23:44 |
| jhesketh | clarkb: actually it's a bash script grabbing the console log | 23:44 |
| jhesketh | the python script is to collate all the required files and make an index and push it up | 23:44 |
| jhesketh | but yeah, it should be doable | 23:44 |
| mikal | So... who wants to talk to me about temptest tests which require modifications to the test guest? | 23:45 |
| jhesketh | clarkb: so the envinject module has been removed from puppet, do we need to delete it off the puppet masters somehow? (I didn't know to express that when I did the change) | 23:45 |
| clarkb | jhesketh: we do | 23:45 |
| clarkb | jhesketh: basically need to do what we do for plugin updates to remove it. Tell jenkins to remove the plugin, restart master | 23:45 |
| mikal | sdague: le ping? | 23:46 |
| clarkb | jhesketh: I may have time to do that tomorrow morning prior to US v Belgium | 23:46 |
| clarkb | jhesketh: but no promises :) | 23:46 |
| jhesketh | mikal is getting ready for the Paris summit clearly | 23:46 |
| sdague | mikal: le pong | 23:46 |
| *** mmaglana has quit IRC | 23:46 | |
| mikal | jhesketh: no, I have a spec where someone wants to add a feature, so I asked about tempest tests | 23:46 |
| sdague | or is it la pong? | 23:46 |
| mikal | They've pointed out that testing in tempest would require modifications to the guest os for the test instance | 23:46 |
| jhesketh | mikal: I was referring to your 'le' ;-) | 23:47 |
| mikal | And I was wondering if we have any precident for that sort of thing | 23:47 |
| mikal | jhesketh: :P | 23:47 |
| mikal | sdague: ^-- all that is the reason for my euro-ping | 23:47 |
| sdague | mikal: explain what needs to be modified. Also lets do this in -qa | 23:47 |
| jhesketh | clarkb: right, I tried to find an example of where that was but I couldn't dig it out of the git tree... any pointers? | 23:47 |
| mikal | sdague: https://review.openstack.org/#/c/72038/ | 23:47 |
| mikal | sdague: sure, moving there | 23:47 |
| clarkb | jhesketh: the process? it may be tribal knowledge >_> | 23:48 |
| *** unicell has quit IRC | 23:48 | |
| clarkb | jhesketh: we should probably update our jenkins doc with a section on plugin modifications | 23:48 |
| clarkb | anteaya: see comment on 102581 | 23:48 |
| *** marcoemorais has quit IRC | 23:49 | |
| *** marcoemorais has joined #openstack-infra | 23:50 | |
| clarkb | mordred: I +2'd 101763 but didn't approve it as I believe that step needs ansible to be working for us and I am not sure if it is fully there yet | 23:50 |
| clarkb | mordred: feel free to approve based on where you are in the ansible stack | 23:50 |
| *** marcoemorais has quit IRC | 23:50 | |
| *** marcoemorais has joined #openstack-infra | 23:50 | |
| *** unicell has joined #openstack-infra | 23:51 | |
| clarkb | AaronGr: hey on 101603 a quick local test has gzip -f foo.gz resulting in foo.gz.gz which ends up being double compressed with a different sha1sum | 23:56 |
| openstackgerrit | A change was merged to openstack-infra/config: Make rally-install-bare-centos6 voting again https://review.openstack.org/102620 | 23:56 |
| clarkb | AaronGr: I am almost positive that this is not what we want as a failover csase | 23:56 |
| clarkb | AaronGr: and the find already excludes .gz files | 23:56 |
| clarkb | AaronGr: is there something I am missing for how this helps? | 23:56 |
| *** devkulkarni has quit IRC | 23:57 | |
| clarkb | jesusaurus: nibalizer ^ do you guys know? | 23:58 |
| *** msabramo has quit IRC | 23:59 | |
| * clarkb walks home but do ping if anyone knows what is going on there | 23:59 | |
| pleia2 | clarkb: hm, good catch | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!