| *** rlandy has quit IRC | 00:31 | |
| *** swest has quit IRC | 01:10 | |
| *** swest has joined #zuul | 01:25 | |
| *** threestrands has quit IRC | 02:10 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool master: Implement a Kubernetes driver https://review.openstack.org/535557 | 02:48 |
|---|---|---|
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/nodepool master: Implement an OpenShift resource provider https://review.openstack.org/570667 | 02:48 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: executor: add support for generic build resource https://review.openstack.org/570668 | 02:49 |
| *** bhavik1 has joined #zuul | 03:12 | |
| *** threestrands has joined #zuul | 03:26 | |
| *** threestrands has quit IRC | 03:26 | |
| *** threestrands has joined #zuul | 03:26 | |
| *** lennyb has quit IRC | 04:08 | |
| *** yolanda has quit IRC | 04:09 | |
| *** lennyb has joined #zuul | 04:22 | |
| *** yolanda has joined #zuul | 04:24 | |
| *** bhavik1 has quit IRC | 04:48 | |
| *** nchakrab has joined #zuul | 05:02 | |
| *** nchakrab_ has joined #zuul | 05:04 | |
| *** nchakrab has quit IRC | 05:07 | |
| *** hwoarang has quit IRC | 05:07 | |
| *** hwoarang has joined #zuul | 05:09 | |
| *** nchakrab has joined #zuul | 05:11 | |
| *** nchakrab_ has quit IRC | 05:15 | |
| *** nchakrab has quit IRC | 05:41 | |
| *** nchakrab has joined #zuul | 05:59 | |
| *** nchakrab_ has joined #zuul | 06:03 | |
| *** nchakrab has quit IRC | 06:07 | |
| *** dtruong2 has joined #zuul | 06:09 | |
| *** dtruong2 has quit IRC | 06:20 | |
| *** pcaruana has joined #zuul | 06:37 | |
| *** hashar has joined #zuul | 07:23 | |
| *** electrofelix has joined #zuul | 07:40 | |
| *** threestrands has quit IRC | 07:45 | |
| *** jpena|off is now known as jpena | 07:53 | |
| *** threestrands has joined #zuul | 08:00 | |
| *** threestrands has quit IRC | 08:03 | |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul-jobs master: Some small improvements to zuul_swift_upload https://review.openstack.org/587345 | 08:49 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul-jobs master: Remove unused code path in zuul_swift_upload https://review.openstack.org/587346 | 08:49 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul-jobs master: Add --dry-run option to zuul_swift_upload https://review.openstack.org/587347 | 08:49 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul-jobs master: Enable options for generating the parent links https://review.openstack.org/587348 | 08:49 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul-jobs master: Allow predefined zuul_log_path during swift upload https://review.openstack.org/587349 | 08:55 |
| tobiash | jhesketh: added a comment on 587345 | 09:00 |
| *** sshnaidm|afk has quit IRC | 09:02 | |
| *** sshnaidm has joined #zuul | 09:04 | |
| *** jimi|ansible has quit IRC | 09:21 | |
| jhesketh | tobiash: thanks, I've repliedc | 09:37 |
| *** panda|rover|off is now known as panda|rover | 09:48 | |
| *** electrofelix has quit IRC | 10:12 | |
| *** hashar is now known as hasharAway | 10:22 | |
| *** electrofelix has joined #zuul | 10:47 | |
| *** swest has quit IRC | 11:07 | |
| *** swest has joined #zuul | 11:21 | |
| *** jesusaur has quit IRC | 11:36 | |
| *** jesusaur has joined #zuul | 11:40 | |
| *** jpena is now known as jpena|lunch | 11:46 | |
| *** tobiash has quit IRC | 12:01 | |
| *** rcarrillocruz has quit IRC | 12:04 | |
| *** nchakrab_ has quit IRC | 12:17 | |
| *** panda|rover is now known as panda|rover|lunc | 12:18 | |
| *** dkranz has joined #zuul | 12:30 | |
| *** nchakrab has joined #zuul | 12:34 | |
| *** rlandy has joined #zuul | 12:35 | |
| *** elyezer has joined #zuul | 12:36 | |
| *** jpena|lunch is now known as jpena | 12:48 | |
| *** samccann has joined #zuul | 12:56 | |
| *** tobiash has joined #zuul | 12:57 | |
| *** dkranz has quit IRC | 13:09 | |
| *** elyezer has quit IRC | 13:27 | |
| *** samccann has quit IRC | 13:31 | |
| *** nchakrab_ has joined #zuul | 13:32 | |
| *** jimi|ansible has joined #zuul | 13:33 | |
| *** jimi|ansible has joined #zuul | 13:33 | |
| *** nchakrab has quit IRC | 13:35 | |
| *** myoung has joined #zuul | 13:37 | |
| *** samccann has joined #zuul | 13:39 | |
| *** elyezer has joined #zuul | 13:40 | |
| *** panda|rover|lunc is now known as panda|rover | 14:01 | |
| *** nchakrab_ has quit IRC | 14:08 | |
| *** nchakrab has joined #zuul | 14:09 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul-jobs master: Enable options for generating the parent links https://review.openstack.org/587348 | 14:14 |
| jhesketh | corvus: heh, I was just adding that whitespace ;-) | 14:14 |
| jhesketh | thanks | 14:14 |
| *** elyezer has quit IRC | 14:15 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul-jobs master: Swift logs: rename POST_RETRIES POST_ATTEMPTS https://review.openstack.org/587479 | 14:15 |
| corvus | jhesketh: oh you're still up! hi! :) | 14:16 |
| corvus | jhesketh, tobiash: ^ i think we can push that whole stack through now | 14:16 |
| jhesketh | corvus: I was just about to fix up the retries thing... do you want me to do that quickly or push through the stack first? | 14:17 |
| *** samccann has quit IRC | 14:17 | |
| corvus | jhesketh: i just renamed it in a followup patch; i say we ask tobiash to +3 the first patch and you to +2 the last one | 14:18 |
| jhesketh | sure | 14:18 |
| tobiash | corvus, mordred: looks like the latest openstacksdk release fixed my USE_SHADE issue :) | 14:18 |
| corvus | tobiash: yep, and our initial tests using that are bearing fruit: https://object-storage-ca-ymq-1.vexxhost.net/swift/v1/86bbbcfa8ad043109d2d7af530225c72/logs_78/587178/1/check/tox-py35/e88b9ea/ | 14:18 |
| corvus | jhesketh: ^ not sure if you saw this in -infra scrollback | 14:19 |
| *** nchakrab has quit IRC | 14:19 | |
| corvus | (probably so if you wrote the "../" patch :) | 14:19 |
| *** nchakrab_ has joined #zuul | 14:19 | |
| pabelanger | re: https://review.openstack.org/584488/ anybody have suggestions how I can best test indefinite launch retries? We don't seem to track retries in zk at all | 14:19 |
| jhesketh | corvus: I hadn't, but it's looking good :-) | 14:19 |
| jhesketh | nah I had picked up on the series of nits I posted when I reviewed the earlier patch, but only just pushed them up | 14:20 |
| corvus | jhesketh: oh i'm sorry, i don't remember seeing those! | 14:20 |
| jhesketh | corvus: no, I didn't add them because they weren't blockers | 14:21 |
| corvus | well, all good points :) | 14:21 |
| jhesketh | we'd spend forever nitpicking over what is a wip tool if I did ;-) | 14:21 |
| jhesketh | I was just a bit slow on the followup sorry | 14:21 |
| corvus | then i have successfully made room for collaboration :) | 14:21 |
| jhesketh | \o/ | 14:22 |
| corvus | pabelanger: maybe you should try a white-box approach -- where you poke at the internal data structures of the requesthandler to find out how many attempts it has tried | 14:23 |
| tobiash | corvus: yay :) | 14:23 |
| *** dkranz has joined #zuul | 14:23 | |
| corvus | pabelanger: (most of our testing is black-box, where we don't rely on knowledge of internal state, but occasionally we have to, and that's okay) | 14:24 |
| tobiash | corvus: I have it in a base-test job in prod too now | 14:24 |
| tobiash | corvus: what I noticed is that the x-delete-after seems to be totally ignored by our ceph-radosgw | 14:24 |
| corvus | mnaser: ^ know anything about that? | 14:25 |
| tobiash | corvus: any objection about adding this info explicitly to the metadata of the object too? | 14:25 |
| corvus | tobiash: i confess, i haven't tried it on vexxhost with a really small value. i guess i should? | 14:25 |
| tobiash | corvus: I know that ceph-radosgs doesn't have the automated cleanup but I hoped that I get this implicitly as metadata | 14:26 |
| tobiash | so I can run a cleanup script periodically | 14:26 |
| corvus | tobiash: that's a thing? oh, the metadata isn't for ceph, it's for you? | 14:26 |
| tobiash | but that doesn't seem to be the case so I think I need to add that timestamp also to the metadata | 14:26 |
| mordred | tobiash: oh yuck - radosgw doesn't honor the expires? :( | 14:26 |
| tobiash | I found no x-delete-* header in the ceph docs | 14:27 |
| tobiash | so I'm pretty sure that there is no autodeleter | 14:27 |
| tobiash | so my way would be to run a maybe nightly job that does the cleanup | 14:28 |
| tobiash | (could be even part of zuul-jobs as well) | 14:28 |
| pabelanger | corvus: sure, let me see if I can access that data | 14:28 |
| corvus | tobiash: can we set arbitrary metadata in the headers when we upload the object? | 14:28 |
| mordred | http://tracker.ceph.com/issues/4099 | 14:28 |
| mordred | fwiw | 14:29 |
| corvus | mordred, tobiash: is it possible it only supports x-delete-at rather than x-delete-after? | 14:29 |
| corvus | that would be super easy to work around | 14:29 |
| corvus | http://tracker.ceph.com/projects/ceph/repository/revisions/4f9a84310e847f10b6c3d0fc573e6dd06bc095c6/diff | 14:30 |
| tobiash | corvus: I think the additional metadata is there as this is really part of the swift api | 14:30 |
| corvus | tobiash: i think it would be fine to add metadata as long as it doesn't add an extra API call. however, can you test whether x-delete-at works first? | 14:31 |
| tobiash | corvus: yes, testing it | 14:31 |
| *** nchakrab has joined #zuul | 14:31 | |
| corvus | (it looks like the GC interval is 10 minutes, based on http://tracker.ceph.com/projects/ceph/repository/revisions/db27ea92494201e127d81498f5918834e04e3d02/diff ) | 14:32 |
| *** samccann has joined #zuul | 14:33 | |
| tobiash | if that would work that would be awesome | 14:34 |
| *** nchakrab_ has quit IRC | 14:35 | |
| *** elyezer has joined #zuul | 14:35 | |
| corvus | hrm, it also looks like it should support x-delete-after http://tracker.ceph.com/projects/ceph/repository/revisions/master/entry/src/rgw/rgw_rest_swift.cc#L767 | 14:38 |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Some small improvements to zuul_swift_upload https://review.openstack.org/587345 | 14:39 |
| mordred | corvus: I don't see that code in the actual ceph repo | 14:39 |
| corvus | http://tracker.ceph.com/projects/ceph/repository/revisions/fa347d8f69b8eff2e246d35a127c4bfa5a50b5e0/diff | 14:39 |
| mordred | corvus: https://github.com/ceph/ceph/search?q=DELETE_AFTER&unscoped_q=DELETE_AFTER | 14:39 |
| tobiash | corvus: hrm, ok I'll retest with just a small number. I expected that I get any info in any header about the expiration date. | 14:39 |
| corvus | mordred: ceph uses submodules, maybe it's in those? | 14:39 |
| corvus | mordred: not a submodule: https://github.com/ceph/ceph/blob/master/src/rgw/rgw_rest_swift.cc | 14:41 |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Remove unused code path in zuul_swift_upload https://review.openstack.org/587346 | 14:41 |
| mordred | corvus: ah - there we go yeah: https://github.com/ceph/ceph/blob/master/src/rgw/rgw_rest_swift.cc#L771 | 14:41 |
| mordred | I was apparently just bad at searching | 14:41 |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Add --dry-run option to zuul_swift_upload https://review.openstack.org/587347 | 14:42 |
| mordred | http://lists.ceph.com/pipermail/ceph-users-ceph.com/2017-June/018367.html | 14:43 |
| corvus | tobiash: are you basing your assertion that x-delete-after doesn't work on having tried it, or just reading the docs, or it not having returned the headers back to you when you performed a GET or HEAD? | 14:43 |
| mordred | at least one year ago someone else was having issues with ceph not expiring things | 14:43 |
| tobiash | corvus: I'm based on reading the docs and getting no info about the expiration date | 14:44 |
| mordred | http://tracker.ceph.com/issues/20177 | 14:44 |
| tobiash | so I'll try it now with 5 minutes and check later if it's still there | 14:44 |
| mordred | " | 14:44 |
| mordred | objects are checked for expiration once per day. The start time for processing is configured by "rgw lifecycle work time", and defaults to midnight." | 14:44 |
| tobiash | mordred: so I have to check tomorrow probably | 14:45 |
| mordred | yes, it seems so | 14:45 |
| corvus | tobiash: in that case, it's worth testing both delete-after and delete-at at the same time on two different objects :) | 14:45 |
| tobiash | corvus: yah, it probably makes sense to do that manually instead of via zuul | 14:46 |
| corvus | tobiash: you can run zuul_swift_upload from the cli; it will use your clouds.yaml file | 14:46 |
| tobiash | corvus: I know, I just deleted that from the workdir an hour ago ;) | 14:47 |
| tobiash | have to set it up again first | 14:47 |
| *** elyezer has quit IRC | 14:48 | |
| *** elyezer has joined #zuul | 14:49 | |
| tobiash | ok, uploaded the test fixtures once with delete-at and once with delete-after... | 14:54 |
| tobiash | let's see tomorrow if that worked | 14:54 |
| * mordred hopes | 14:54 | |
| tobiash | oh, looks like delete-after is already gone | 14:55 |
| tobiash | delete-at now too | 14:55 |
| corvus | \o/ easy resolution! | 14:55 |
| mordred | yay! | 14:56 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul-jobs master: Allow predefined zuul_log_path during swift upload https://review.openstack.org/587349 | 14:57 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul-jobs master: Add delete_after to swift upload role https://review.openstack.org/587493 | 14:57 |
| tobiash | so we should plumb that into a role var ^ | 14:57 |
| *** nchakrab_ has joined #zuul | 14:59 | |
| *** nchakrab has quit IRC | 15:02 | |
| *** pcaruana has quit IRC | 15:10 | |
| *** nchakrab_ has quit IRC | 15:32 | |
| *** sshnaidm is now known as sshnaidm|off | 15:36 | |
| *** elyezer has quit IRC | 15:36 | |
| *** elyezer has joined #zuul | 15:57 | |
| mnaser | corvus: it should work, did it work with a small value? | 15:59 |
| mnaser | it is supported afaik and there is even tempest tests which test against that | 15:59 |
| mnaser | did it end up working? | 16:00 |
| clarkb | mnaser: tobiash's tests were successful | 16:03 |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Enable options for generating the parent links https://review.openstack.org/587348 | 16:04 |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Swift logs: rename POST_RETRIES POST_ATTEMPTS https://review.openstack.org/587479 | 16:11 |
| *** myoung is now known as myoung|lunch | 16:24 | |
| *** sdake has quit IRC | 16:33 | |
| corvus | mnaser: yep, sorry, false alarm! :) | 16:33 |
| *** sdake has joined #zuul | 16:33 | |
| *** sdake has joined #zuul | 16:33 | |
| *** dtruong2 has joined #zuul | 16:36 | |
| *** sshnaidm|off has quit IRC | 16:43 | |
| *** panda|rover is now known as panda|rover|off | 16:49 | |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Allow predefined zuul_log_path during swift upload https://review.openstack.org/587349 | 17:02 |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Add delete_after to swift upload role https://review.openstack.org/587493 | 17:03 |
| *** jpena is now known as jpena|off | 17:05 | |
| corvus | mordred: i don't see that swift has support for symlinks. there is a trick you can do with manifests to have an object return the content of a different existing object, which sort of approximates symlinked files, but i don't think that's going to work for symlinked directories (unless you also did it for every file in the target directory). | 17:07 |
| corvus | mordred: given that the use case you're presenting is a recursive symlink ( symlink -> . ), that may not be the best idea. we could decide to do it for one level deep and stop, i guess. but that's starting to feel weird. | 17:08 |
| corvus | mordred: i wonder if we should just change that to a copy instead of a symlink in the post playbook? | 17:09 |
| corvus | clarkb: are there other symlink use cases we should consider? anything in devstack you know of? | 17:09 |
| clarkb | I think following symlinks and copying the "result" is probablu fine | 17:09 |
| corvus | clarkb: for files only, or also directories? | 17:09 |
| clarkb | I'm not sure if yhere are specific cases of one or the other | 17:10 |
| clarkb | but if foo -> bar/ and bar/ has A and B in it we can copy those as foo/A foo/B ? | 17:10 |
| corvus | yeah, we could do that. that works for files and directories. what about recursive directory symlinks though? copy one level or ignore them? | 17:11 |
| clarkb | foo -> baz -bar/ ? | 17:11 |
| clarkb | hrm | 17:12 |
| corvus | (mutually recursive directory symlinks also need to be considered, of course: a->b b->a) | 17:12 |
| corvus | (projecting symlinks onto a system that doesn't support them is hard) | 17:13 |
| clarkb | foo -> baz -> bar/ would still be foo/A and foo/B I think | 17:14 |
| clarkb | foo -> baz/C -> bar becomes foo/C/A ? | 17:14 |
| clarkb | or do we flatten flattening may be the easiest consistent thing to do | 17:15 |
| clarkb | I want to say that is how os.walk returns them too | 17:15 |
| tobiash | https://specs.openstack.org/openstack/swift-specs/specs/in_progress/symlinks.html | 17:16 |
| tobiash | That's probably unimplemented? | 17:16 |
| corvus | i'm assuming so. i can't find any reference to it in the docs | 17:17 |
| tobiash | Is there a need for symlinks? | 17:18 |
| mordred | tobiash: zuul-build-dashboard-multi-tenant uses them | 17:19 |
| corvus | tobiash: we have one use case in the zuul-build-dashboard-multi-tenant job. but i think that can be replaced with copying files. i'm unsure if there are others. | 17:19 |
| mordred | corvus: I agree re: copying | 17:19 |
| clarkb | corvus: I think loops we ignore (just delete them or add a file at the first "entrypoint" to loop that says "this is a symlink loop, not copying") then for everything else do what os.walk does and return files relative to the logs/ dir and not their abspaths | 17:20 |
| clarkb | that way the tree looks as presented to zuul when uploaded | 17:20 |
| corvus | clarkb: os.walk doesn't descend into symlink targets which are directories | 17:21 |
| clarkb | corvus: my local testing shows it follows them | 17:24 |
| tobiash | It can but it cannot detect loops | 17:24 |
| corvus | By default, walk() will not walk down into symbolic links that resolve to directories. Set followlinks to True to visit directories pointed to by symlinks, on systems that support them. | 17:24 |
| corvus | https://docs.python.org/2/library/os.html | 17:24 |
| clarkb | yup, I'm saying we should do the followlinks=True behavior then check for loops like zuul does | 17:25 |
| corvus | clarkb: it will return the symlink as a directory, but won't descend into it | 17:25 |
| clarkb | http://paste.openstack.org/show/726961/ we would copy ./testing_dir/bar2 and ./foo/baz/bar2 | 17:26 |
| corvus | clarkb: so basically copy the result, unless we end up looping, in which case we just stop descending? | 17:26 |
| clarkb | corvus: ya I think if we detect a loop we can add a file with "symlink loop detected" at that point | 17:26 |
| clarkb | and stop looping | 17:26 |
| mordred | yeah - and in zuul-build-dashboard-multi-tenant - we can switch the multi.yaml playbook to do copying itself instead of creating symlinks | 17:27 |
| corvus | okay. i agree that's sensible. however, i don't believe we have any current use cases which would benefit from that, so i'm tempted to leave that for someone else to implement. :) | 17:28 |
| corvus | i can write that up as a TODO | 17:29 |
| corvus | clarkb: where's the os.walk loop detection in zuul you were thinking of? | 17:31 |
| *** gchenuet has joined #zuul | 17:32 | |
| corvus | clarkb: oh, it's the ansible path stuff. got it | 17:32 |
| clarkb | corvus: zuul.ansible.paths._is_safe_path() | 17:32 |
| corvus | okay, i'll just implement this. the TODO comment is nearly as long as the implementation anyway. :) | 17:32 |
| gchenuet | Hi guys ! I'm working on TLS and Gearman with ansible-role-zule. But i faced some `[SSL: WRONG_VERSION_NUMBER] wrong version number` on gearman server log. | 17:33 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Use copy instead of symlink for multi-tenant dashboard https://review.openstack.org/587554 | 17:33 |
| openstackgerrit | Paul Belanger proposed openstack-infra/zuul-jobs master: Add nodepool info to test-emit-job-header https://review.openstack.org/557947 | 17:33 |
| gchenuet | I'm using self-signed cert | 17:33 |
| gchenuet | Am i missing something ? | 17:33 |
| pabelanger | clarkb: corvus: mordred: a little late, but updated ^ to add more nodepool information for emit-job-header | 17:34 |
| *** myoung|lunch is now known as myoung | 17:34 | |
| pabelanger | gchenuet: it should work, we use SSL cert for gate jobs in windmill. | 17:35 |
| pabelanger | gchenuet: https://git.zuul-ci.org/cgit/zuul/tree/tests/fixtures/gearman/README.rst might help | 17:36 |
| gchenuet | should i need to used the openstack-gearman-role to setup gearman or only the zuul-role is good enough ? | 17:37 |
| gchenuet | thanks for the link ! | 17:38 |
| gchenuet | We used the Gearman SSL page to generate our cert: http://gearman.info/gearmand/ssl.html | 17:38 |
| pabelanger | don't need ansible-role-gearman, that is only if you want a stand alone gearman server | 17:38 |
| gchenuet | ok ! I'm trying your link ! | 17:39 |
| gchenuet | thanks a lot ! | 17:39 |
| pabelanger | make sure you have gearman_server.start=true in zuul.conf | 17:39 |
| pabelanger | along with proper SSL certs | 17:39 |
| *** gchenuet has quit IRC | 17:40 | |
| corvus | it's more complicated than the code that's in the ansible stuff (since it needs to actually handle recursion). i'm going to leave a note and call it done. | 17:47 |
| corvus | (the ansible path checks short-circuits if it's seen a directory already, but it's not a stack, it's just a list, so if we applied the same algorithm, we'd get different results based on ordering) | 17:49 |
| clarkb | corvus: we would still short circuit though right? just maybe not at the same location every job run? | 17:51 |
| corvus | clarkb: yeah. we'd end up randomly picking one directory to upload if it is the target of multiple symlinks | 17:53 |
| clarkb | I think that sort is stable based on locale fwiw | 17:53 |
| corvus | /s/randomly/arbitrarily/ | 17:54 |
| clarkb | but ya that won't fix the case of mordred needing multiple copies of the same thing | 17:56 |
| mordred | clarkb: https://review.openstack.org/587554 has fix for that | 17:57 |
| clarkb | for the logs case I think that behavior is sufficient for the artifacts case it isn't. hrm | 17:59 |
| clarkb | the arbitrary selection not mordreds fix | 17:59 |
| corvus | clarkb: what's the artifacts case? | 17:59 |
| openstackgerrit | Paul Belanger proposed openstack-infra/zuul-jobs master: Add nodepool info to test-emit-job-header https://review.openstack.org/557947 | 18:00 |
| clarkb | corvus: one example is mordreds symlink case I think, its the same files shared across multiple dirs for different tenants? but also say you tag a release and build docs for that but also want to update the latest docs publications you might symlink in the job to accomplish that | 18:00 |
| clarkb | corvus: all of these cases can be solved job side by using a copy or hardlink instead of a symlink | 18:00 |
| *** electrofelix has quit IRC | 18:07 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul-jobs master: Swift logs: don't allow links outside of the supplied path https://review.openstack.org/587580 | 18:19 |
| corvus | clarkb, mordred, tobiash: ^ that closes a symlink vulnerability and adds a test for the current behavior | 18:19 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul-jobs master: Swift logs: don't allow links outside of the supplied path https://review.openstack.org/587580 | 18:20 |
| *** dkranz has quit IRC | 20:03 | |
| *** samccann has quit IRC | 20:28 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul-jobs master: Compress svg graphics during swift upload https://review.openstack.org/587616 | 20:43 |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: web: add tenant-scoped autohold, enqueue (change, ref) https://review.openstack.org/576907 | 20:44 |
| tobiash | corvus: I have some big svg images (various system stats) and compressing would change them from 1.1M to 75K, so I suggest to also compress svg images ^ | 20:45 |
| corvus | ++ | 21:02 |
| *** yolanda has quit IRC | 21:25 | |
| *** myoung has quit IRC | 21:28 | |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: web: add tenant-scoped, JWT-protected actions https://review.openstack.org/576907 | 21:34 |
| openstackgerrit | Matthieu Huin proposed openstack-infra/zuul master: web: add tenant-scoped, JWT-protected actions https://review.openstack.org/576907 | 21:58 |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Compress svg graphics during swift upload https://review.openstack.org/587616 | 22:55 |
| *** hasharAway has quit IRC | 23:06 | |
| *** threestrands has joined #zuul | 23:24 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!