2023-11-02 08:02:54 <ncopa> morning
2023-11-02 08:03:00 <ncopa> mkdir: can't create directory '/builds/alpine/aports/community/openmp/src': No space left on device
2023-11-02 08:03:04 <ncopa> CI is out of diskspace
2023-11-02 08:03:47 <mps> sorry, I made a mess with !54607
2023-11-02 08:04:07 <mps> and I have no idea how to fix it
2023-11-02 08:04:43 <ncopa> oh.... haha!
2023-11-02 08:04:46 <ncopa> no worries
2023-11-02 08:05:04 <mps> (now I'm thinking this gitlab workflow is not for me, and I should stop to work with it)
2023-11-02 08:05:38 <ncopa> no, dont worry
2023-11-02 08:05:51 <ncopa> what happened was you had wrong target branch
2023-11-02 08:06:01 <mps> yes
2023-11-02 08:06:08 <ncopa> i changed it from master to 3.16-stable now
2023-11-02 08:06:49 <ikke> ncopa: the runners on usa-bld-1 have limited disk space 
2023-11-02 08:08:29 <ncopa> i also think that dotnet is building?
2023-11-02 08:08:47 <ncopa> which eats all diskspace it gets, regardless how much disk you have...
2023-11-02 08:09:09 <mps> patchwwork was a better workflow for me
2023-11-02 08:09:31 <ncopa> i think you do pretty well with gitlab
2023-11-02 08:10:30 <ncopa> and im happy that you are able to adapt. patchwork was a mess to keep track of what was fixed and what was not
2023-11-02 08:10:36 <ncopa> and was a mess when it came to rebases etc
2023-11-02 08:10:45 <mps> could be, but I always feel some 'fear in fingers' when creating or merging something
2023-11-02 08:10:45 <ikke> And no CI 
2023-11-02 08:11:35 <ncopa> a little "fear" may be healthy. makes you be careful.
2023-11-02 08:11:41 <mps> yes, I remember that I proposed switch from patchwork to gitlab instead to github
2023-11-02 08:11:56 <ncopa> and im very happy we did that
2023-11-02 08:12:28 <mps> yes, yes, overall it is really good for alpine
2023-11-02 08:12:37 <ncopa> and dont worry. messing up a little bit happens to all of us
2023-11-02 08:13:00 <ncopa> thankfully there are protections in gitlab to prevent real disasters (like force push to master  branch)
2023-11-02 08:14:01 <mps> and we have to give big thanks to ikke and clandmeter for making gitlab working very well
2023-11-02 08:19:13 <mps> btw, I noticed that some uses telegram-desktop, and it is lagging in alpine. I have it fixed for few months with this APKBUILD locally https://tpaste.us/ggYx if someone need new and fancy version
2023-11-02 08:21:52 <ncopa> seems like we have issues with s390x again?
2023-11-02 08:35:47 <ikke> ncopa: i checked this morning, and it still continued 
2023-11-02 08:36:17 <ikke> Just a large llvm rebuild that's taking a long time 
2023-11-02 14:11:29 <nmeum> was go bootstrap for 3.19 fixed? I was on vacation last week and just came back. didn't look into it in the meantime
2023-11-02 14:11:34 <nmeum> *on i386
2023-11-02 14:12:04 <ikke> no, it still fails
2023-11-02 14:12:38 <ikke> https://build.alpinelinux.org/buildlogs/build-3-19-x86/community/go/go-1.21.3-r1.log
2023-11-02 14:20:44 <nmeum> ok
2023-11-02 16:12:29 <ncopa> clandmeter, we need to renew the docker sponsored open source program each year. its just a formality
2023-11-02 16:12:38 <ncopa> ..said justin
2023-11-02 16:13:04 <ncopa> who redirected me to the correct person. I need to fill in the form again
2023-11-02 16:13:09 <ikke> Ok, how do we do that?
2023-11-02 16:13:16 <ikke> ok
2023-11-02 16:13:56 <ncopa> https://www.docker.com/community/open-source/application/
2023-11-02 16:14:09 <ncopa> I'm working on it, but I'm not sure how we filled it in last time
2023-11-02 16:14:56 <ikke> It's something you did. You just mentioned that you did it
2023-11-02 16:16:36 <ncopa> ha!
2023-11-02 16:19:22 <ncopa> Enter the existing Docker ID for your organization on Docker Hub.
2023-11-02 16:19:28 <ncopa> I suppose that is alpinelinux, right?
2023-11-02 16:19:32 <ikke> correct
2023-11-02 16:19:44 <ncopa> https://hub.docker.com/u/alpinelinux
2023-11-02 16:19:45 <ikke> I mean, that's the namespace
2023-11-02 16:19:51 <ikke> indeed
2023-11-02 16:26:18 <ncopa> How many core developers are actively contributing to this project?
2023-11-02 16:26:25 <ncopa> those questions are hard :) 
2023-11-02 16:26:37 <ikke> yes
2023-11-02 16:26:45 <ikke> 2-4
2023-11-02 16:31:14 <ncopa> do we have an infra@alpinelinux.org email address?
2023-11-02 16:31:26 <ncopa> they send notifications to ncopa@alpinelinux.org currently
2023-11-02 16:31:38 <ncopa> which likely drown in my inbox
2023-11-02 16:34:53 <ikke> yes, we have
2023-11-02 16:35:40 <ncopa> it is infra@alpinelinux.org?
2023-11-02 16:36:02 <ikke> yes
2023-11-02 16:38:26 <ncopa> I have sent an owner invite to that email
2023-11-02 16:40:32 <ikke> Ok, that means we need to create an account with that e-mail?
2023-11-02 16:43:11 <ikke> Created one
2023-11-02 19:49:48 <noleak238974> hello
2023-11-02 19:49:56 <noleak238974> whats up guys
2023-11-02 19:53:09 <ikke> mps: someone in #alpine-linux asks about https://gitlab.alpinelinux.org/alpine/aports/-/issues/14476
2023-11-02 19:54:04 <mps> ikke: iirc I answered this question long ago
2023-11-02 19:55:05 <ikke> At least that ticket does not have a response and I could not find another ticket
2023-11-02 19:55:35 <mps> well, maybe not answered but asked for reasonable explanation why this is needed. Telling 'it is enabled on -lts' is not good enough reason imo
2023-11-02 19:57:03 <mps> but maybe 'we' should enable it? would be nice to hear opinion of more devs
2023-11-02 19:57:40 <noleak238974> guys i have a question
2023-11-02 19:58:38 <mps> noleak238974: don't ask to ask on IRC, just ask
2023-11-02 19:59:18 <noleak238974> okay okay tx can i use anonsurf from parrot linux in alpine linux
2023-11-02 20:01:14 <mps> you should ask this question on #alpine-linux channel, which is for user help
2023-11-02 20:01:41 <mps> is parrot linux musl or glibc based
2023-11-02 20:02:26 <noleak238974> i think it´s glibc but i m not sure
2023-11-02 20:03:25 <mps> if it is glibc based then binaries from it probably will not work on alpine
2023-11-03 08:11:07 <clandmeter> tomalok: what is your email address?
2023-11-03 08:11:20 <clandmeter> maybe i can get it from gitlab
2023-11-03 08:14:08 <clandmeter> tomalok, ikke: i created an agenda item with google meet. should popup in your mailbox.
2023-11-03 08:15:13 <ikke> Received it 
2023-11-03 11:28:29 <ikke> ncopa: Does the suggestion in https://gitlab.alpinelinux.org/alpine/security/secfixes-tracker/-/merge_requests/16 make sense? I'd like to merge and deploy that MR
2023-11-03 11:35:36 <ncopa> not sure. there is something current_app from flask i think: https://flask.palletsprojects.com/en/2.3.x/appcontext/#purpose-of-the-context
2023-11-03 11:35:44 <ncopa> how can i reproduce the issue?
2023-11-03 11:36:56 <ikke> I ran it with: https://gitlab.alpinelinux.org/alpine/infra/docker/secfixes-tracker and then this diff https://tpaste.us/VqD8
2023-11-03 11:37:19 <ikke> docker compose run -u python --entrypoint /bin/sh crond
2023-11-03 11:38:55 <ikke> ncopa: " View functions, error handlers, and other functions that run during a request will have access to current_app."
2023-11-03 11:41:16 <ikke> In the container, run flask update-states (not sure if you need to run something like import-nvd recent first)
2023-11-03 11:45:22 <ncopa> i wonder if we can catch it in the testsuite?
2023-11-03 11:45:30 <ncopa> the error
2023-11-03 11:46:28 <ikke> In the function tests you run the importers, can you run update-states similarly as well?
2023-11-03 11:49:12 <ncopa> ah, i havent done the update-states tests yet?
2023-11-03 11:49:38 <ikke> no
2023-11-03 11:50:12 <ncopa> that explains
2023-11-03 11:50:29 <ncopa> iirc the udpate-states needed a bit more work than the other tests
2023-11-03 11:56:14 <ncopa> looks like i already discussed it with chatgpt some time ago
2023-11-03 11:58:20 <ikke> problem is the current_app is not available in that context (I don't think there are global variables that automatically are made available)
2023-11-03 12:18:10 <ncopa> maybe we can use db.get_app()  https://flask-sqlalchemy.palletsprojects.com/en/2.x/api/#flask_sqlalchemy.SQLAlchemy.get_app
2023-11-03 12:18:30 <ncopa> -        return self.package_name in current_app.config.get('PACKAGE_EXCLUSIONS', [])
2023-11-03 12:18:30 <ncopa> +        return self.package_name in db.get_app().config.get('PACKAGE_EXCLUSIONS', [])
2023-11-03 12:23:05 <ikke> 👍
2023-11-03 12:47:05 <ncopa> nope. does not work they removed get_app() in SQLAlchemy 3.x
2023-11-03 12:48:15 <ikke> Ok
2023-11-03 12:51:21 <ncopa> I don't think adding a global is a good solution, but I don't know how to fix it properly
2023-11-03 12:51:41 <ikke> Yeah, I'm not a fan of that myself either
2023-11-03 12:52:33 <ncopa> but I have a test for it
2023-11-03 12:52:47 <ikke> I'm not sure the db model is the correct place for that logic either 
2023-11-03 12:54:58 <ncopa> ah... we only need to import current_app
2023-11-04 21:04:19 <ikke> cleaned up usa2 a bit. distfiles took up some space
2023-11-06 10:42:01 <mps> !54890
2023-11-06 11:00:47 <mps> fix '// typedef Elf64_Relr GElf_Relr;' in /usr/include/gelf.h pkg elfutils-dev
2023-11-06 11:01:01 <mps> is omni here?
2023-11-06 11:10:18 <mps> 'typedef Elf64_Relr GElf_Relr' is probably bug
2023-11-06 11:23:27 <ncopa> im working on fixing musl for Elf64_Relr
2023-11-06 18:23:15 <mps> ncopa: now that linux-asahi kernel works fine with GPU I think to merge mesa-asahi to testing. any objection?
2023-11-06 18:50:56 <mps> do anyone have some time to review !54926
2023-11-06 19:40:31 <ncopa> do we need a separate build for asahi?
2023-11-06 19:48:42 <mps> ncopa: yes, mainline mesa doesn't work with asahi GPU
2023-11-06 19:51:10 <mps> asahi project also have it https://github.com/AsahiLinux/PKGBUILDs/tree/main/mesa-asahi-edge
2023-11-06 19:51:23 <mps> and I think fedora, debian ...
2023-11-06 19:51:53 <ptrc> ikke, ncopa or someone else with builder access: could you check why lld16 didn't get built on edge aarch64?
2023-11-06 19:52:11 <ptrc> or, well, did get built but not uploaded to the mirrors(?)
2023-11-06 19:56:51 <mps> did I made pkgver wrong? If not I think it is ready to merge
2023-11-06 19:57:23 <mps> someone with better knowledge about mesa could fix and polish it later
2023-11-06 19:58:02 <mps> I use it for about 10-11 months now, iirc
2023-11-06 19:58:32 <mps> and some other people also by downloading from my repo
2023-11-06 20:00:49 <ptrc> mps: if you mean the mesa-asahi MR, imo the pkgver looks correct
2023-11-06 20:01:10 <mps> ptrc: yes, I mean it. thanks
2023-11-06 20:01:22 <ptrc> the commented out deps and source could be removed, but that's just a nitpick
2023-11-06 20:01:59 <mps> yes, I know it is not proper APKBUILD but that could be fixed later
2023-11-06 20:03:16 <mps> after nearly two years working on making alpine good on apple silicon I think it is time to put all in aports so other people can make it better
2023-11-06 20:03:47 <mps> and it will be easier for end users to install and use
2023-11-06 20:04:47 <mps> I prepared bottle of good red wine to celebrate, hehe
2023-11-06 20:18:05 <ncopa> ptrc: lld16 was built on edge aarch64
2023-11-06 20:18:59 <ncopa> and it was uploaded: https://dl-master.alpinelinux.org/alpine/edge/community/aarch64/
2023-11-06 20:19:42 <ptrc> ah, i've checked a while ago and it wasn't there yet but the buildlog was, so i assumed something went wrong
2023-11-06 20:20:04 <ptrc> but yeah, i can see it now in the apkindex as well
2023-11-06 20:20:07 <ptrc> thanks
2023-11-06 20:23:54 <mps> ok, merged mesa-asahi
2023-11-06 20:29:14 <mps> and tested this from alpine repo, work very well.
2023-11-06 20:29:17 <mps> \o/
2023-11-06 20:30:06 <mps> after two years alpine on apple silicon have all needed pkgs in aports
2023-11-06 20:31:05 <mps> though alpine was first distro used as daily driver on apple silicon, by tmlind (I only know his IRC nick) and me
2023-11-06 20:31:23 <mps> now wine, cheers!!!
2023-11-06 20:37:13 <ikke> ptrc: aarch64 edge was blocked by qt6-qtwebengine for some time
2023-11-06 20:38:55 <mps> ikke: it is disabled by omni on aarch64 and I disabled calibre because it depends on qt6-qtwebengine
2023-11-06 20:39:43 <ikke> Yes, I'm aware, just stating why lld16 was not available for some time
2023-11-07 21:33:45 <ikke> ncopa: starting secfixes-tracker locally consumes a lot of memory
2023-11-07 21:33:58 <ikke> Immediately on startup it takes ~8GB
2023-11-07 21:35:10 <ncopa> hum, ok
2023-11-07 21:35:20 <ncopa> just starting it?
2023-11-07 21:35:25 <ncopa> or importing stuff
2023-11-07 21:36:34 <ikke> just starting it
2023-11-07 21:42:33 <ikke> trying to get a memory profile, but takes forever
2023-11-07 21:48:58 <ikke> ncopa: seems to also happen with 0.4.0, so not a new symptom
2023-11-07 21:53:58 <ikke> doesn't seem to happen with 0.3.1
2023-11-07 21:58:20 <ncopa> i can have a look at the log tomorrow
2023-11-07 22:00:19 <ncopa> what about 0.3.2?
2023-11-07 22:01:38 <ncopa> and 0.3.3
2023-11-08 06:24:53 <ikke> ncopa: 0.3.3 low memory, 0.4.0 high memory
2023-11-08 14:02:52 <ncopa> so its likely due to my performance optimizations when importing stuff
2023-11-08 14:07:06 <ikke> I was bisecting it
2023-11-08 16:22:34 <ikke> ncopa: seems to be the difference between alpine 3.15 and 3.18
2023-11-08 16:22:46 <ikke> (didn't check any versions in between yet)
2023-11-08 16:23:26 <ikke> sorry, no it was not running yet
2023-11-08 16:32:28 <ikke> It's strange, it's doing it even on older versions
2023-11-08 16:32:39 <ikke> and it's a sudden increase in memory usage, not gradual
2023-11-08 16:37:38 <ncopa> is it the public site?
2023-11-08 16:38:32 <ikke> yes
2023-11-08 16:38:39 <ikke> running it in the docker compose project
2023-11-08 16:38:50 <ncopa> so it may be accessed from outside?
2023-11-08 16:39:05 <ikke> Oh sorry, this one is just running locally on my machine
2023-11-08 16:39:11 <ncopa> ok
2023-11-08 16:39:23 <ncopa> is it a cron job that does imports?
2023-11-08 16:39:48 <ikke> Yes, but I'm not running the cron service
2023-11-08 16:40:01 <ncopa> maybe the data it uses during imports is not garbage collected?
2023-11-08 16:40:15 <ikke> It's the uwsgi process that is taking the memory
2023-11-08 16:41:15 <ikke> python         9  5.8 51.3 8452204 8382712 ?     S    16:40   0:02 uwsgi --master --plugin python3 --manage-script-name --mount /=secfixes_tracker:app --plugin http --http 0.0.0.0:8080 --processes=1 --threads=1
2023-11-08 16:41:21 <ikke> using 8G memory
2023-11-08 16:41:27 <ikke> immediately after startup
2023-11-08 16:41:48 <ncopa> do you have much in the database?
2023-11-08 16:42:00 <ikke> Starting with a database from scratch
2023-11-08 16:42:12 <ncopa> so an empty db?
2023-11-08 16:42:57 <ikke> yes
2023-11-08 16:45:43 <ikke> Note that I don't see it happening in production, so it may be something local
2023-11-08 16:46:27 <ncopa> from memory_profiler import profile
2023-11-08 16:46:35 <ncopa> have you tried memory_profiler?
2023-11-08 16:47:19 <ikke> no, not yet
2023-11-08 18:44:44 <ikke> ncopa: memory usage seems normal when running flask in development mode, so I think it has to do with uwsgi
2023-11-08 18:51:20 <ikke> I think I may have found it
2023-11-08 18:54:37 <ikke> ncopa: "detected max file descriptor number: 1073741816"
2023-11-08 18:55:04 <ikke> I lowered it to 2048 and memory usage is normal
2023-11-08 19:00:13 <ikke> https://github.com/unbit/uwsgi/issues/2299
2023-11-08 19:11:30 <ncopa> interesting
2023-11-08 19:11:53 <ikke> yes
2023-11-08 19:12:10 <ncopa> good job finding it!
2023-11-08 19:12:34 <ikke> I just happen to notice in the output of running uwsgi
2023-11-09 06:44:13 <ikke> cleaned up usa-bld-1
2023-11-09 07:15:39 <ncopa> thanks
2023-11-09 10:12:50 <PureTryOut> ncopa: linux-lts currently fails to build on x86_64 builder due to lack of space
2023-11-09 10:14:47 <ncopa> mkdir: can't create directory '/home/buildozer/aports/main/linux-lts/src': No space left on device
2023-11-09 10:24:42 <ncopa> slow machine...
2023-11-09 11:38:53 <ikke> ncopa: are you cleaning up nld9, or should I do that?
2023-11-09 11:39:26 <ncopa> i cleaned a bit already
2023-11-09 11:39:39 <ncopa> but please go ahead and try clean it up even more if possible
2023-11-09 11:39:45 <ikke> ok, will do
2023-11-09 11:39:55 <ncopa> i wonder if we should start remove older builders
2023-11-09 11:40:14 <ncopa> i think it was full because there were several failed builds, which left the sources trees there
2023-11-09 11:40:15 <ikke> Yeah, would make sense, but at the same time, the older builders are relativelt small
2023-11-09 11:40:23 <ikke> ncopa: yeah, definitely
2023-11-09 11:40:34 <ikke> dotnet, chromium, qt6-qtwebengine, etc
2023-11-09 11:40:37 <ncopa> yup
2023-11-09 11:40:54 <ncopa> and there were multiple kernel buidls at the same time
2023-11-09 11:40:57 <ikke> We need at some point think about community
2023-11-09 11:41:07 <ikke> it's growing immensely
2023-11-09 11:41:53 <ncopa> yes
2023-11-09 11:45:00 <ncopa> [9564446.734442] oom-kill:constraint=CONSTRAINT_NONE,nodemask=(null),cpuset=lxc.payload.build-3-9-x86_64,mems_allowed=0-1,global_oom,task_memcg=/lxc.payload.build-3-19-x86_64,task=regression_test,pid=26809,uid=1000
2023-11-09 11:45:09 <ncopa> it runs out of memory too...
2023-11-09 11:46:37 <ikke> hmm
2023-11-09 11:48:18 <ikke> didn't realize linux also takes a lot of space to build
2023-11-09 11:48:31 <ikke> 21G source dir
2023-11-09 13:03:06 <ikke> ncopa: 40G .cargo dir on build-edge-x86_64 
2023-11-09 13:03:11 <ikke> pruning it now
2023-11-09 13:04:47 <ncopa> great! thanks!
2023-11-09 13:05:08 <ikke> deployed secfixes-tracker v0.4.1 
2023-11-09 13:11:26 <ikke> ncdu is a great tool to clean up disk space
2023-11-09 14:15:00 <raspbeguy> yes it is
2023-11-09 14:48:21 <ikke> clandmeter: fyi, I'm running garbage collection on our docker registry to clean it up a bit
2023-11-09 16:23:22 <ikke> clandmeter: Object storage usage went from ~70G to ~15G :)
2023-11-09 16:24:30 <ikke> hope nothing broke, initial tests seems things are still working
2023-11-09 19:23:29 <clandmeter> nice
2023-11-09 19:24:59 <ikke> Gives me a bit more confidence in the sustainability
2023-11-09 19:27:50 <ikke> I had one issue where one image couldn't be pushed, but apparently that was some bug in the container registries s3 implementation, found a post on the linode community forum with the solution
2023-11-09 19:29:05 <ikke> clandmeter: for simple cases, it should be enough to include https://gitlab.alpinelinux.org/alpine/infra/gitlab-ci-templates/-/blob/master/exec/docker-image-x86_64.yml in the .gitlab-ci.yml file to push an image to the registry
2023-11-10 13:50:20 <mps> aports/main/librtlsdr needs some care
2023-11-10 13:51:22 <mps> I changed it a lot locally, not sure if it is ok to take maintainership, me or anyone else active
2023-11-11 20:11:49 <mps> algitbot: retry master
2023-11-12 07:41:58 <ikke> kunkku: do we manually need to renew spoke certificates for dmvpn? I just hade one site with an expired certificate
2023-11-12 07:45:17 <ikke> There is also a hub-1 cert that has expired
2023-11-12 09:20:20 <mps> how to find what block 'apk upgrade'? apk fix doesn't help
2023-11-12 11:20:18 <ikke> mps: block in what way?
2023-11-12 11:21:30 <mps> ikke: `apk upgrade` doesn't work, just says 'OK: 3543 MiB in 1157 packages' but nothing is upgraded
2023-11-12 11:22:00 <mps> I guess something in dependencies block it but can't find what
2023-11-12 11:22:37 <mps> `apk fix -v` says same
2023-11-12 11:23:01 <ikke> yeah, apk fix does not do anything in that regard
2023-11-12 11:23:11 <ikke> mps: what does apk version -l '<' return?
2023-11-12 11:23:56 <mps> show pkgs which should be upgraded
2023-11-12 11:24:24 <mps> same as `apk version` only
2023-11-12 11:24:27 <ikke> and then `apk add -s pkgname=version`
2023-11-12 11:24:40 <ikke> where you try to explicitly install a newer version of a package
2023-11-12 11:24:48 <ikke> oh, and apk list --orphan
2023-11-12 11:27:21 <mps> unrar, vim-help, svt-av1-libs, light and some locally build pkgs
2023-11-12 11:27:35 <ikke> from the last command?
2023-11-12 11:27:42 <mps> yes
2023-11-12 11:27:52 <ikke> Either one of those could be blocking other packages from being upgraded
2023-11-12 11:29:16 <mps> here is output https://tpaste.us/BJ5W
2023-11-12 11:29:35 <mps> output of the `apk add -s firefox=119.0.1-r0`
2023-11-12 11:30:26 <ikke> what arch?
2023-11-12 11:31:07 <ikke> apk info -R unrar vim-help <other-packages>
2023-11-12 11:31:35 <mps> uhm, a lot of manual works
2023-11-12 11:31:58 <mps> icu-data is culprit
2023-11-12 11:32:36 <ikke> Somethign is keeping the old version of icu-data-en installed
2023-11-12 11:33:15 <mps> current solution for is `apk del fluffychat`
2023-11-12 11:34:08 <mps> price for running edge
2023-11-12 11:34:13 <ikke> yes
2023-11-12 11:34:34 <mps> lets hope fluffychat will be rebuilt soon
2023-11-12 11:34:49 <ikke> what arch?
2023-11-12 11:35:11 <mps> aarch64
2023-11-12 11:35:18 <ikke> right, testing still needs to be completed
2023-11-12 11:35:35 <mps> and libreoffice with upgrade right now is not fixed
2023-11-12 11:35:37 <ikke> It's building electron now
2023-11-12 11:35:58 <mps> yes, I follow build procces 
2023-11-12 11:36:41 <mps> this is how I'm 'testing' edge
2023-11-12 11:37:06 <ikke> yes, which is good, but after big rebuilds like icu things are broken for a while
2023-11-12 11:37:25 <mps> though I would really like if we have less pkgs
2023-11-12 11:39:23 <ikke> The problem is that everyone has a different set of minimal amount of packages they need
2023-11-12 11:41:12 <mps> I thought busybox is enough for working system ;p
2023-11-12 22:14:15 <ptrc> https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/55177
2023-11-12 22:14:21 <ptrc> ikke: suspiciously stuck pipeline
2023-11-13 20:30:48 <mps> I bet a lot of people will ask soon for bcachefs-tools in alpine, especially when we upgrade linux-edge to 6.7
2023-11-13 20:31:47 <mps> I made pkg and thinking to merge it to testing for now, but maybe even to community where is linux-edge
2023-11-13 20:33:00 <mps> have to add that these days I don't have much free time to test it extensively
2023-11-14 00:18:59 <ptrc> ikke: do we have some stuff left over on the 3.19 ppc64le builder? i'm trying to reproduce the iproute2 failure locally and it doesn't install some of the stuff that the builder does, for some reason
2023-11-14 00:19:07 <ptrc> https://tpaste.us/vNLW
2023-11-14 05:04:15 <ikke> ptrc: no, world is clean
2023-11-14 06:30:16 <ikke> ptrc: these are the packages that are installed in a clean docker container installing the depends and makepends from iproute2: https://tpaste.us/QRk8 
2023-11-14 06:32:40 <ikke> ptrc: note that I did check the elf headers on ppc64le and they did not include the missing symbol / macro
2023-11-15 11:15:01 <mps> https://gitlab.alpinelinux.org/alpine/aports/-/jobs/1182389#L71
2023-11-15 11:15:30 <mps> anyone know why it can't find rdfind? I works fine in my LXC
2023-11-15 14:48:44 <mps> ncopa: next longterm stable kernel is 6.6, https://www.kernel.org/category/releases.html
2023-11-15 14:49:35 <mps> so !54530 should be merged
2023-11-15 16:06:36 <ncopa> mps: cool. I'm having an alpine break for two weeks. will get to it when I get back
2023-11-15 16:07:19 <ncopa> ikke, clandmeter, docker ppl reached out to us an mentioned that someone is using alpine namespace: https://hub.docker.com/u/alpine. Their github repositories are here: https://github.com/alpine-docker
2023-11-15 16:07:37 <mps> ncopa: I can merge it if you agree
2023-11-15 16:08:01 <ncopa> I confirmed that it is not us or official alpine. they asked If we wanted them to reclaim the namespace for us?
2023-11-15 16:08:08 <ncopa> mps: i will look at it when I get back
2023-11-15 16:08:13 <mps> ok
2023-11-15 16:08:17 <ikke> We were aware of that, but never took any action against it
2023-11-15 16:08:26 <ncopa> they ask if we want take action
2023-11-15 16:08:37 <mps> so alpine 3.19 will be later
2023-11-15 16:08:39 <ncopa> I wonder if we should ask them to join us and make it official
2023-11-15 16:08:39 <ikke> Yeah, the name can imply it's sanctioned / provided by us
2023-11-15 16:09:06 <ikke> How would that look like
2023-11-15 16:09:12 <ncopa> i dont know
2023-11-15 16:09:14 <ikke> Not sure if we want to disrupt the current images
2023-11-15 16:09:21 <ncopa> exactly
2023-11-15 16:09:29 <ncopa> and I don't know if we care too much either
2023-11-15 16:09:44 <ncopa> they did not use the official alpine logo, so it seems like they want to play nice
2023-11-15 16:09:48 <ncopa> but I don't know
2023-11-15 16:10:16 <ncopa> so i respond to docker that we don't bother to take action yet, but maybe later?
2023-11-15 16:10:42 <ncopa> it is misleading
2023-11-15 16:11:38 <ncopa> ikke: do you think you can ask around in the team what alpine people would prefer?
2023-11-15 16:11:47 <ikke> ncopa: sure
2023-11-15 16:11:47 <ncopa> I'm slightly busy with other stuff right now
2023-11-15 16:11:51 <ncopa> thanks!
2023-11-15 16:11:51 <ikke> yeah, no worry
2023-11-15 22:01:26 <kunkku> ikke: about renewing dmvpn certs
2023-11-15 22:02:33 <kunkku> 'dmvpn-ca cert generate' renews all certs that have expired or will soon expire
2023-11-15 22:03:39 <kunkku> setup-dmvpn must be run on the respective hosts with the new pfx file
2023-11-16 01:36:54 <ncopa> mps: i was about to look at the linux 6.6 thing but I realised that you talked about the headers
2023-11-16 01:37:38 <ncopa> the headers is only for compiling user space, and we already built 90+ of user space for alpine 3.19
2023-11-16 01:38:21 <ncopa> if we upgrade the headers, we risk that things may break when rebuilds (eg when we do security fixes for the future 3.19-stable branch)
2023-11-16 01:38:42 <ncopa> so I think we should wait with linux headers 6.6 til after the 3.19 release
2023-11-16 01:38:59 <ncopa> the headers themselves does not really provide any value
2023-11-16 01:39:48 <ncopa> what we may consider is bump linux-its to 6.6
2023-11-16 01:40:18 <ncopa> but I don't want include more work at this point
2023-11-16 11:31:18 <mps> ncopa: yes, I talked about linux-headers, not linux-lts
2023-11-16 11:32:24 <mps> I can't remember when upgrading linux-headers break anything
2023-11-16 11:32:46 <mps> most safe pkg in alpine
2023-11-16 11:32:54 <mps> safest*
2023-11-19 16:56:09 <ikke> clandmeter: I've tagged _rc1, but the trigger failed, says that there is no permission to trigger downstream pipeline
2023-11-19 16:58:50 <clandmeter> Yes
2023-11-19 16:59:04 <clandmeter>  you are not ncopa 
2023-11-19 16:59:27 <clandmeter> I added him to the project 
2023-11-19 16:59:41 <clandmeter> I’m in Spain atm
2023-11-19 16:59:57 <clandmeter> can you add yourself and try retry?
2023-11-19 17:00:11 <clandmeter> Best is to move the project
2023-11-19 17:12:39 <ikke> clandmeter: yeah, works after I've added myself
2023-11-19 17:40:58 <clandmeter> Something changed in gitlab in recent versions that broke it.
2023-11-19 17:41:43 <ikke> They tightened the implicit permissions
2023-11-19 17:43:04 <ikke> https://gitlab.alpinelinux.org/clandmeter/alpine-disk-image/-/settings/ci_cd Token access section
2023-11-19 20:47:34 <tomalok> ikke, clandmeter: rc1 successfully tagged?  i can start doing tome cloud image testing...
2023-11-19 20:47:54 <ikke> tomalok: yes
2023-11-19 20:48:19 <ikke> tomalok: I'll try to work next week on the image forwarder
2023-11-19 20:48:37 <tomalok> are we expecting a release of the 3.16..3.18 around the same time as 3.19?
2023-11-19 20:48:57 <tomalok> ikke: 👍 i've got some signing work to do yet too
2023-11-19 20:49:11 <ikke> tomalok: I don't know when ncopa expects to do those minor releases
2023-11-19 20:49:36 <tomalok> saw something about openssl in devel
2023-11-19 20:49:39 <ikke> yeah
2023-11-19 20:57:47 <tomalok> fwiw it's 3.19_rc1 and not 3.19.0_rc1? (thought we had a 'z' last time)
2023-11-19 20:58:31 <tomalok> either that or my code to resolve the latest release is off
2023-11-19 20:59:39 <tomalok> tag is right. it's releases.json maybe?
2023-11-19 20:59:54 <ikke> no, alpine-base :(
2023-11-19 21:03:16 <tomalok> apk_ver = self.apk_version('main', 'x86_64', 'alpine-base', ver=ver) yep, looks like that's where it comes from
2023-11-19 21:03:18 <ikke> I can fix the package, but that won't change what's in the image
2023-11-19 21:03:53 <tomalok> let me see how it got resolved in the work/images.yaml -- version vs release
2023-11-19 21:04:13 <ikke> real fix would be rc2
2023-11-19 21:04:37 <tomalok>   release: 3.19_rc1
2023-11-19 21:04:37 <tomalok>   version: '3.19'
2023-11-19 21:04:52 <tomalok> ThisIsFine™
2023-11-19 21:05:11 <tomalok> especially since this won't be published
2023-11-19 21:05:21 <tomalok> or released
2023-11-19 21:07:53 <ikke> Ok
2023-11-19 23:13:33 <tomalok> 3.19_rc1 aws images all seem fine (x86_64/aarch64, bios/uefi, tiny/cloudinit)
2023-11-22 08:11:06 <clandmeter> good morning
2023-11-22 08:11:14 <clandmeter> ikke: thx for fixing things
2023-11-22 08:11:23 <clandmeter> i did not expect you to tag releases :)
2023-11-22 08:19:06 <ikke> I tagged an rc while ncopa was away (with his approval)
2023-11-22 11:47:06 <ikke> I rebooted it
2023-11-22 11:51:12 <ikke> FYI I've stopped the 3.19 arm* builders so that edge can build qt5-qtwebengine. Otherwise it runs OOM
2023-11-22 14:38:05 <mps> is the riscv64 builder stuck?
2023-11-23 14:42:54 <mps> is it to late to add new openssl 3.2 to alpine 3.19 https://www.openssl.org/blog/blog/2023/11/23/OpenSSL32/
2023-11-23 14:44:07 <ikke> yes
2023-11-23 14:46:07 <mps> yes means late?
2023-11-23 14:46:13 <ikke> indeed
2023-11-23 14:47:15 <mps> it's a pity, but ...
2023-11-23 14:56:34 <ikke> all of 3.19 has already been built, and we want to make a release. Rebuilding against a new openssl would cost a lot of time and would perhaps encounter new build issues
2023-11-23 14:59:24 <mps> I know I know
2023-11-27 07:26:52 <ncopa> are we having issues with the s390x machine? im trying to upgrade my s390x dev container, but its dog slow. and CI appears to be broken for s390x
2023-11-27 07:26:59 <ncopa> https://gitlab.alpinelinux.org/alpine/aports/-/jobs/1194624
2023-11-27 07:34:59 <ncopa> usa2-dev1 [~]# uptime
2023-11-27 07:34:59 <ncopa>  07:34:42 up 47 days, 18:59,  load average: 3.46, 2.88, 3.80
2023-11-27 07:35:26 <ncopa> what is causing the load avg to be that high?
2023-11-27 07:55:10 <ncopa> gitlab-runner-s390x is also unusually slow. im tempted to reboot it
2023-11-27 07:55:44 <ncopa> maybe I should try upgrade the kernel and reboot it
2023-11-27 08:00:18 <ikke> Don't see a constant high load on that machine
2023-11-27 08:42:34 <ncopa> i rebooted it
2023-11-27 12:35:51 <ncopa> Is che-bld-1 unavailable for everyone or just me?
2023-11-27 12:50:19 <ikke> ncopa: i can ssh into it 
2023-11-27 13:23:35 <ncopa> aha, its only my tunnels that are broke
2023-11-27 13:44:24 <ncopa> my dmvpn is broken. AUTH_FAILED
2023-11-27 13:44:27 <ncopa> Nov 27 13:43:03 rtr-copa daemon.info : 09[NET] received packet: from 172.105.69.172[500] to 85.167.243.86[500] (61 bytes)
2023-11-27 13:44:27 <ncopa> Nov 27 13:43:03 rtr-copa daemon.info : 05[ENC] parsed IKE_AUTH response 1 [ N(AUTH_FAILED) ]
2023-11-27 13:44:27 <ncopa> Nov 27 13:43:03 rtr-copa daemon.info : 09[IKE] received AUTHENTICATION_FAILED notify error
2023-11-27 13:44:27 <ncopa> Nov 27 13:43:03 rtr-copa daemon.info : 09[ENC] parsed IKE_AUTH response 1 [ N(AUTH_FAILED) ]
2023-11-27 13:44:27 <ncopa> Nov 27 13:43:03 rtr-copa daemon.info : 05[NET] received packet: from 176.58.106.16[500] to 85.167.243.86[500] (61 bytes)
2023-11-27 13:44:29 <ncopa> Nov 27 13:43:03 rtr-copa daemon.info : 05[IKE] received AUTHENTICATION_FAILED notify error
2023-11-27 14:40:49 <clandmeter> cert expired?
2023-11-27 15:13:48 <ncopa> could be. not sure how I check that
2023-11-27 15:14:29 <ikke> /etc/swanctl
2023-11-27 15:14:40 <ikke> Or on dmvpn1.a.o
2023-11-27 15:40:02 <ikke> OU=DEV-NCOPA, CN=VPNc-1   11/15/18 10:32:53  Expired 11/14/23 10:32:53
2023-11-27 15:40:05 <ikke> ncopa: 
2023-11-27 16:15:32 <ikke> so yes, it's expired
2023-11-27 16:33:43 <ncopa> thank you for helping finding that out. I'll create a new one (later)
2023-11-27 16:33:56 <ncopa> banging my head into this rpi stuff
2023-11-28 08:45:41 <mps> ncopa: what you think, should we move some asahi pkg to community
2023-11-28 08:51:00 <ncopa> good question. maybe we should
2023-11-28 08:53:15 <mps> m1n1, u-boot-asahi, tiny-dfr and maybe asahi-fwextract are candidates
2023-11-28 11:18:37 <ikke> ncopa: when you have some time, could you take a look at https://gitlab.alpinelinux.org/alpine/security/secfixes-tracker/-/merge_requests/17? This will make sure we're ready for the decomissioning of the NVD feeds at the 15th
2023-11-28 11:25:47 <ncopa> ok will try have a look at that today, but I need to prioritize (rpi) kernels
2023-11-28 11:26:08 <ikke> Yes, no hurry
2023-11-28 13:10:58 <ncopa> drats... che-bld-1 is missing and I had some work inprogress there...
2023-11-28 13:11:46 <ncopa> i can continue to work on my mbp but i dont know if I should redo the work or wait for the che machine to come back
2023-11-28 13:20:36 <mps> ncopa: looks like it works now
2023-11-28 13:21:12 <mps> hmm, no
2023-11-28 13:21:28 <mps> how then I work on my lxc there
2023-11-28 13:46:47 <ikke> ncopa: seems resolved
2023-11-28 13:46:57 <ncopa> yup
2023-11-28 13:47:14 <ncopa> im working on pushing it to the MR so I can move my work out from there
2023-11-28 15:54:13 <ncopa> ikke: how do I rebase an MR which uses 'master' as branch?
2023-11-28 15:54:20 <ncopa> eg https://gitlab.alpinelinux.org/alpine/mkinitfs/-/merge_requests/131
2023-11-28 15:56:56 <ikke> You need to unprotect it in the fork
2023-11-28 15:57:24 <ikke> In the repository settings of the fork 
2023-11-28 16:00:23 <ncopa> got it. thanks!
2023-11-29 16:34:58 <ptrc> ikke: any chance i could get access to some aarch64 machine? trying to debug this firefox issue locally but it takes ages to rebuild everything
2023-11-29 16:35:10 <ikke> ptrc: yeah
2023-11-29 16:38:56 <ptrc> how would i go about that then? :p
2023-11-29 16:39:39 <ikke> do you have ipv6 connectivity?
2023-11-29 16:39:54 <ptrc> yeah, i think so
2023-11-29 17:12:39 <ikke> guess I'll have to clean up again
2023-11-29 17:19:59 <ncopa> thank you so much for taking care of it
2023-11-30 02:35:41 <ayakael0> Btw, MilkV is releasing the following this summer: https://community.milkv.io/t/introducing-the-milk-v-oasis-with-sg2380-a-revolutionary-risc-v-desktop-experience/780
2023-11-30 02:36:32 <ayakael0> 16 core RISCV64 cpu for 160 USD, seems too good to be true given the pioneer's 1500 USD price tag
2023-11-30 02:37:16 <ayakael0> None the less,4-5 of those in a cluster might do something nice, although one beefy system would be better for compilation services
2023-11-30 09:26:09 <ncopa> i think I may need generate a new cert for my dmvpn, but I dont remember how or where
2023-11-30 09:37:12 <ncopa> ok, i am not able to log in to build-3-17-armhf. I cannot tag release until i know whats going on there.
2023-11-30 09:37:23 <ncopa> apparently my ssh key is not on usa9-dev1.a.o either
2023-11-30 09:41:29 <clandmeter> its on the master iirc
2023-11-30 09:41:41 <ncopa> the dmvpn.a.o?
2023-11-30 09:41:47 <clandmeter> one of them yes
2023-11-30 09:42:11 <ncopa> ah! we have two of them!
2023-11-30 09:42:26 <clandmeter> high availability :)
2023-11-30 09:43:31 <ikke> dmvpn1
2023-11-30 09:43:47 <ncopa> yup i found it. trying to figure out how to generate a cert
2023-11-30 09:44:11 <clandmeter> i guess you could use shell history
2023-11-30 09:44:16 <clandmeter> that would be my first go to :)
2023-11-30 09:44:46 <clandmeter> and having the password would be a prio 
2023-11-30 10:01:41 <ikke> dmvpn-ca cert generate site <name>
2023-11-30 11:18:42 <ncopa> now to figure out how to install the cert
2023-11-30 11:19:17 <ikke> dmvpn-setup <cert>
2023-11-30 11:19:33 <ncopa> does that work even if dmvpn is already setup?
2023-11-30 11:20:03 <ikke> yes
2023-11-30 11:20:18 <ikke> That's how I updated the cert on usa2
2023-11-30 11:21:26 <ncopa> setup-dmvpn
2023-11-30 11:21:30 <ncopa> but yay! that worked
2023-11-30 11:57:02 <ncopa> https://security.alpinelinux.org/vuln/CVE-2023-5678 shows possible vulnerable for 3.15, 3.16, 3.17, but i think we have backported the patch for those, and its in secfixes comment.
2023-11-30 11:57:13 <ncopa> 109910c17d51d060eec185e649bb6a2d513f010d
2023-11-30 11:57:26 <ncopa> 3.16: 109910c17d51d060eec185e649bb6a2d513f010d
2023-11-30 11:58:07 <ncopa> 3.17: I added it a minut ago, so that may show up later. 36ea41b5179ee066fc011b52b63040d900e0b872
2023-11-30 11:59:07 <ncopa> 3.15: afb7a0b731d6486b1dd51c531f7783dae3f1f9a3