| *** gtristan has quit IRC | 05:42 | |
| *** gtristan has joined #baserock | 06:05 | |
| *** toscalix has joined #baserock | 07:18 | |
| *** paulwaters_ has joined #baserock | 07:51 | |
| *** rdale has joined #baserock | 08:01 | |
| *** jonathanmaw has joined #baserock | 08:48 | |
| *** locallycompact has joined #baserock | 08:59 | |
| *** locallycompact has joined #baserock | 09:00 | |
| *** ssam2 has joined #baserock | 09:07 | |
| *** ChanServ sets mode: +v ssam2 | 09:07 | |
| *** adds68 has quit IRC | 09:22 | |
| *** adds68 has joined #baserock | 09:22 | |
| *** toscalix has quit IRC | 10:08 | |
| *** toscalix has joined #baserock | 10:25 | |
| gtristan | ssam2, jjardon... can you point me to the toolchain thing that you are currently converting with ? | 10:30 |
|---|---|---|
| gtristan | So I can do the arches migration ? | 10:30 |
| ssam2 | it's the elements/gnu-toolchain stack in definitions.git master | 10:30 |
| ssam2 | the conversion is done from definitions.git master, wrapped by a script named 'convert' | 10:31 |
| gtristan | definitions.git master... somewhere on gitlab.com I think | 10:31 |
| gtristan | not easily findable on the wiki, maybe good to have in the topic line here ? | 10:31 |
| ssam2 | https://gitlab.com/baserock/definitions | 10:31 |
| ssam2 | the wiki links to Gitlab on the front page, but yeah there is a lot of cruft | 10:32 |
| gtristan | nice, so submit MR and will use latest buildstream master ? | 10:32 |
| ssam2 | yeah should all work | 10:32 |
| * gtristan means, to see the CI | 10:32 | |
| ssam2 | ah except the CI is broken | 10:32 |
| ssam2 | kinda | 10:33 |
| ssam2 | i tried to make caching work, and somehow made things even worse | 10:33 |
| ssam2 | builds start at least | 10:33 |
| gtristan | ssam2, but we get passed launching buildstream at least ? | 10:33 |
| gtristan | at least we'll know that the thing loads without the arches conditionals | 10:33 |
| ssam2 | ah yeah it'll get that far | 10:33 |
| ssam2 | it'll just probably run out of disk space at some point | 10:34 |
| gtristan | okay ! so I better not do this all at once... | 10:41 |
| gtristan | ssam2, looks like the base-runtime and base-sdk in definitions is still using the old, misguided thing | 10:42 |
| gtristan | because I didnt know what the hell I was doing back then | 10:42 |
| ssam2 | right, there's a merge request to rip that out | 10:42 |
| gtristan | can we merge it ? | 10:42 |
| ssam2 | we could | 10:43 |
| ssam2 | https://gitlab.com/baserock/definitions/merge_requests/58 | 10:43 |
| ssam2 | of course the CI doesn't pass because the CI doesn't work properly because i broke it | 10:43 |
| ssam2 | oh, actually it's also blocked on https://gitlab.com/baserock/definitions/merge_requests/59/commits | 10:43 |
| ssam2 | which we could merge | 10:43 |
| ssam2 | but i basically ran out of time to finish this work off, and am tied up in other stuff now | 10:43 |
| ssam2 | nobody is using the buildstream conversions, so i'm happy to be cavalier about merging stuff to it | 10:44 |
| gtristan | Ah | 10:45 |
| gtristan | Ok, we aught to get the CI running again though :-/ | 10:46 |
| jjardon | ssam2: what is broken? Is there a bug somewhere with the details? Maybe I can spend some time on the weekend | 10:46 |
| ssam2 | https://gitlab.com/baserock/definitions/merge_requests/62 | 10:47 |
| ssam2 | somewhere a job failed with out of disk space, trying to find that | 10:47 |
| ssam2 | https://gitlab.com/baserock/definitions/-/jobs/35099426 | 10:48 |
| ssam2 | the point I got up to was... i realised that if we store stuff in /cache, gitlab CI will just ignore it because it's outside the build directory | 10:48 |
| jjardon | 8h20m0s -> is normal that it takes this long? | 10:48 |
| gtristan | jjardon, It seems the cache is somehow broken | 10:48 |
| ssam2 | that's the problem I was trying to solve | 10:48 |
| gtristan | or not big enough | 10:48 |
| ssam2 | which is that there was no caching of sources | 10:48 |
| ssam2 | and never actually has been, even for YBD | 10:48 |
| ssam2 | because everything was stored in /cache, which gitlab CI ignores | 10:49 |
| ssam2 | it will have worked by chance when consecutive jobs execute on the same runner, but that's it | 10:49 |
| ssam2 | so I tried to move the /cache volume to /builds, so we'd get enough disk space to store all the cache stuff in the build directory | 10:49 |
| ssam2 | so that gitlab CI would actually upload it to the minio cache we set up | 10:50 |
| ssam2 | which is now working | 10:50 |
| ssam2 | but ... https://gitlab.com/baserock/definitions/-/jobs/35099426 shows gitlab CI running out of space *in the build dir* while preparing a .zip file to upload to the cache | 10:50 |
| jjardon | you can add [ci skip] to the commit message and builds will be skipped, in case nobody have time to fix this | 10:51 |
| jjardon | ssam2: gtristan ^ | 10:51 |
| ssam2 | might help | 10:51 |
| * gtristan was actually hoping to *test* the arches change... | 10:57 | |
| benbrown_ | jjardon: where's the remote cache setup, ooi? | 10:57 |
| jjardon | sec | 10:58 |
| jjardon | benbrown_: cant access from here, but if you go to digital ocean, the machine that says "cache" in the name | 10:59 |
| ssam2 | 46.101.48.48 | 10:59 |
| benbrown_ | ta | 10:59 |
| ssam2 | it's in a docker container on there | 10:59 |
| ssam2 | the minio cache is exposed on port 9005 | 10:59 |
| ssam2 | so you can do `ssh -L9000:localhost:9005 ...` and then browse to https://localhost:9000/ | 11:00 |
| ssam2 | you need an access code that can be found in `docker logs minio` | 11:00 |
| gtristan | ssam2, shall we also change 'bst-arch' and 'bst-target-arch' thingies in the conversion script to be 'arch', 'target-arch' and such ? | 11:00 |
| gtristan | I think so | 11:01 |
| ssam2 | sounds good to me | 11:01 |
| gtristan | Ok so | 11:01 |
| gtristan | I think what I'm gonna do, is do it on my machine | 11:01 |
| gtristan | ssam2, but you wanted that other bug to land first right ? | 11:01 |
| gtristan | fixes for "other arches" ? | 11:01 |
| ssam2 | yeah that's pretty needed for armv8 and ppc64 to work | 11:02 |
| ssam2 | so makes sense to base your work off that | 11:02 |
| ssam2 | it's had months of testing, it's just CI blocking it | 11:02 |
| ssam2 | i suggest basing off https://gitlab.com/baserock/definitions/merge_requests/58 too | 11:02 |
| gtristan | mkay, I'll locally setup a branch with those applied | 11:03 |
| gtristan | then make my changes, which are local to the conversion stuff anyway | 11:03 |
| gtristan | And make sure it builds | 11:03 |
| gtristan | And then monday we can take a last look and just push it | 11:03 |
| paulsherwood | /win 12 | 11:04 |
| paulsherwood | gaaah | 11:04 |
| gtristan | win win win ! | 11:04 |
| paulsherwood | i'm trying :) | 11:05 |
| gtristan | me too ! | 11:21 |
| benbrown_ | ssam2: There are a couple of runners that have been around for 9+ days, were you using these? Or should I nukify them? | 11:32 |
| paulsherwood | do you hve to ask? :) | 11:33 |
| benbrown_ | I'm unsure whether or not they were being used to get a picture of the filesystem | 11:34 |
| * benbrown_ shall assume anything with "runner" in it's name that's this old needs to be destroyed | 11:35 | |
| benbrown_ | s/it\'/it/ | 11:35 |
| paulsherwood | i would, yes | 11:36 |
| gtristan | ssam2, do we have a canonical list, with your patch MR 59 applied... of arches that are supported ? | 11:41 |
| gtristan | ssam2, also, how does this work with elements/gnu-toolchain.bst ? | 11:41 |
| * gtristan a bit confused, that only contains intel arches | 11:41 | |
| ssam2 | benbrown_, they would be useful for debugging probably | 11:42 |
| ssam2 | did you delete already ? | 11:42 |
| benbrown_ | ssam2: Any idea where this should be looking? https://gitlab.com/baserock/definitions/-/jobs/36142506 | 11:42 |
| ssam2 | that's a new one | 11:42 |
| benbrown_ | Or should it be cloning buildstream instead of expecting it to be at a particular path? | 11:42 |
| benbrown_ | ssam2: I did already delete | 11:43 |
| ssam2 | ah | 11:43 |
| ssam2 | it should be cloning buildstream, and already does AFAIK | 11:43 |
| benbrown_ | I can't see anything in the yaml that does that | 11:43 |
| ssam2 | in which branch? | 11:43 |
| ssam2 | hmmm... i wonder if a line got deleted something | 11:44 |
| benbrown_ | I'm working off sam/fix-ci, should I not be? | 11:44 |
| ssam2 | hmm, seems in master that it doesn't actually clone buildstream | 11:44 |
| ssam2 | weird | 11:45 |
| ssam2 | ok, so probably the docker image changed underneath | 11:45 |
| ssam2 | still, the stanza beginning "# Update to latest BuildStream" was intended to actually update to latest buildstream | 11:45 |
| ssam2 | oh wait I know what i did | 11:46 |
| ssam2 | and this is indeed very dumb | 11:46 |
| ssam2 | but was relying on the docker image containing a git clone of buildstream.git in /root | 11:46 |
| ssam2 | which it now doesn't | 11:46 |
| ssam2 | to be honest, you could remove that whole stanza since the docker image is updated nightly anyway and contains buildstream | 11:46 |
| benbrown_ | oh, ok | 11:46 |
| * benbrown_ shall do that then | 11:46 | |
| *** gtristan has quit IRC | 11:52 | |
| *** CTtpollard has quit IRC | 12:00 | |
| *** CTtpollard has joined #baserock | 12:05 | |
| gitlab-br-bot | infrastructure: merge request (sam/readme-in-gitlab->master: Rename README so it gets displayed in GitLab) #10 changed state ("opened"): https://gitlab.com/baserock/infrastructure/merge_requests/10 | 12:12 |
| gitlab-br-bot | infrastructure: merge request (sam/readme-in-gitlab->master: Rename README so it gets displayed in GitLab) #10 changed state ("opened"): https://gitlab.com/baserock/infrastructure/merge_requests/10 | 12:13 |
| gitlab-br-bot | infrastructure: merge request (sam/readme-in-gitlab->master: Rename README so it gets displayed in GitLab) #10 changed state ("merged"): https://gitlab.com/baserock/infrastructure/merge_requests/10 | 12:13 |
| ssam2 | thanks! | 12:13 |
| *** gtristan has joined #baserock | 12:47 | |
| gitlab-br-bot | infrastructure: merge request (sam/update-ca-certs->master: Update CA certs) #11 changed state ("opened"): https://gitlab.com/baserock/infrastructure/merge_requests/11 | 12:48 |
| ssam2 | baserock SSL certificates should be valid for another 3 months now | 12:48 |
| ssam2 | i've shut doiwn the openid, storyboard and gerrit services | 12:50 |
| ssam2 | *shut down | 12:50 |
| ssam2 | as promised in https://listmaster.pepperfish.net/pipermail/baserock-dev-baserock.org/2017-September/013812.html | 12:50 |
| jjardon | ssam2: cool, thanks | 12:51 |
| gitlab-br-bot | definitions: merge request (staging/jjardon/ostree-upgrade->master: strata/ostree-core.morph: Update ostree to 2016.15) #24 changed state ("closed"): https://gitlab.com/baserock/definitions/merge_requests/24 | 12:54 |
| jjardon | mmm, this can be useful: https://www.digitalocean.com/community/tutorials/an-introduction-to-digitalocean-spaces | 13:00 |
| jjardon | benbrown_: ^ | 13:00 |
| *** toscalix has quit IRC | 13:10 | |
| benbrown_ | jjardon: interesting | 13:10 |
| benbrown_ | jjardon: What're you running on that cache instance that gives us s3 like access? | 13:11 |
| jjardon | benbrown_: minions | 13:12 |
| jjardon | :) | 13:12 |
| jjardon | https://github.com/minio/minio | 13:12 |
| benbrown_ | oic | 13:13 |
| benbrown_ | "A large ecosystem of S3-compatible tools and libraries can be used to manage your Space." | 13:15 |
| gitlab-br-bot | ybd: merge request (benbrown/exit-on-multiple-rpms->master: Exit should multiple files match the prebuilt rpm pattern) #392 changed state ("opened"): https://gitlab.com/baserock/ybd/merge_requests/392 | 13:22 |
| gitlab-br-bot | ybd: merge request (benbrown/exit-on-multiple-rpms->master: Exit should multiple files match the prebuilt rpm pattern) #392 changed state ("opened"): https://gitlab.com/baserock/ybd/merge_requests/392 | 13:25 |
| gitlab-br-bot | ybd: merge request (benbrown/exit-on-multiple-rpms->master: Exit should multiple files match the prebuilt rpm pattern) #392 changed state ("merged"): https://gitlab.com/baserock/ybd/merge_requests/392 | 13:25 |
| gitlab-br-bot | definitions: merge request (sam/fix-ci->master: Fix caching in CI) #62 changed state ("opened"): https://gitlab.com/baserock/definitions/merge_requests/62 | 13:40 |
| benbrown_ | jjardon: fyi, I have a branch: benbrown/sam/fix-ci | 13:40 |
| *** CTtpollard has quit IRC | 13:41 | |
| *** CTtpollard has joined #baserock | 13:42 | |
| gitlab-br-bot | definitions: merge request (sam/fix-ci->master: Fix caching in CI) #62 changed state ("opened"): https://gitlab.com/baserock/definitions/merge_requests/62 | 13:49 |
| gitlab-br-bot | definitions: merge request (sam/fix-ci->master: Fix caching in CI) #62 changed state ("opened"): https://gitlab.com/baserock/definitions/merge_requests/62 | 13:50 |
| benbrown_ | jjardon: D: | 13:53 |
| jjardon | benbrown_: uh, seems you branch is doing more thing than simply fixing the CI; I will stop now | 13:53 |
| benbrown_ | tries to optimise stuff as well, sure | 13:54 |
| benbrown_ | Given we're limited by space, it made sense to not pull in both ybd and bst caches into all jobs | 13:55 |
| benbrown_ | And I think we saw some issues with artifacts and dependencies as well, so I've made sure we're explicit about which jobs have dependencies | 13:56 |
| *** toscalix has joined #baserock | 13:56 | |
| *** toscalix has quit IRC | 13:56 | |
| *** toscalix has joined #baserock | 13:57 | |
| jjardon | yeah, nice | 13:57 |
| jjardon | my two little patches should be enough to actually fix the current CI, I think. But yeah better to apply your branch | 13:59 |
| jjardon | mmm, maybe not | 14:00 |
| gitlab-br-bot | definitions: merge request (sam/fix-ci->master: Fix caching in CI) #62 changed state ("opened"): https://gitlab.com/baserock/definitions/merge_requests/62 | 14:12 |
| gitlab-br-bot | ybd: merge request (benbrown/old-git-sha-contains->master: Use branch in place of for-each-ref) #393 changed state ("opened"): https://gitlab.com/baserock/ybd/merge_requests/393 | 15:06 |
| *** locallycompact has quit IRC | 15:13 | |
| *** toscalix has quit IRC | 15:23 | |
| *** CTtpollard has quit IRC | 15:28 | |
| gitlab-br-bot | ybd: merge request (benbrown/old-git-sha-contains->master: Use branch in place of for-each-ref) #393 changed state ("opened"): https://gitlab.com/baserock/ybd/merge_requests/393 | 15:40 |
| gitlab-br-bot | ybd: merge request (benbrown/old-git-sha-contains->master: Use branch in place of for-each-ref) #393 changed state ("merged"): https://gitlab.com/baserock/ybd/merge_requests/393 | 15:40 |
| gitlab-br-bot | definitions: merge request (benbrown/sam/fix-ci->master: Fix caching in CI and some other optimizations) #63 changed state ("opened"): https://gitlab.com/baserock/definitions/merge_requests/63 | 16:21 |
| gitlab-br-bot | definitions: merge request (sam/fix-ci->master: Fix caching in CI) #62 changed state ("closed"): https://gitlab.com/baserock/definitions/merge_requests/62 | 16:22 |
| *** tiagogomes has quit IRC | 16:22 | |
| *** jonathanmaw has quit IRC | 16:35 | |
| *** ssam2 has quit IRC | 16:57 | |
| *** jude_ has quit IRC | 17:08 | |
| *** jude_ has joined #baserock | 18:34 | |
| *** jude_ has quit IRC | 20:12 | |
| *** jude_ has joined #baserock | 20:22 | |
| *** jude_ has quit IRC | 20:29 | |
| *** rdale has quit IRC | 21:31 | |
| *** jude_ has joined #baserock | 22:21 | |
| *** jude_ has quit IRC | 22:54 | |
| *** jude_ has joined #baserock | 23:08 | |
| *** jude_ has quit IRC | 23:12 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!