archlinux-docker issueshttps://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues2022-04-13T16:06:49Zhttps://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/65Arch in Docker as a daily driver2022-04-13T16:06:49ZThomas Buckley-HoustonArch in Docker as a daily driverI've been using this image for a few years now as a daily driver on a Digital Ocean VPS. I do this partly because DO no longer natively support Arch, but also just because it's more convenient to migrate my setup to a new machine by just...I've been using this image for a few years now as a daily driver on a Digital Ocean VPS. I do this partly because DO no longer natively support Arch, but also just because it's more convenient to migrate my setup to a new machine by just copying a Docker image. By "daily driver" I mean I use it for writing code, Matrix chat clients, Syncthing backups, etc.
I've noticed that the setup for this isn't well documented online. Am I really such a minority use case? As such I've taken the initiative of documenting my own approach here: https://github.com/tombh/arch-in-docker-for-daily-use I wonder if anybody here could take a look? I don't know if this is an appropriate place to post this? Apologies if not. But if there are any obvious improvements I'd love to get advice. Thank you very much.https://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/64Bogus entries in default PATH2022-04-13T14:27:48ZAntonio RojasBogus entries in default PATHThe default PATH in a container is
```
# echo $PATH
/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin
```
`/usr/sbin`, `/sbin` and `/bin` should not be there: they are symlinks to `/usr/bin` and having them in PATH is known to...The default PATH in a container is
```
# echo $PATH
/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin
```
`/usr/sbin`, `/sbin` and `/bin` should not be there: they are symlinks to `/usr/bin` and having them in PATH is known to cause issues with cmake and autotools (they don't resolve symlinks so they detect binaries in the symlinked path, then they try to set other paths relative to that which fails).https://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/63xz to produce base.tar.xz fails in low mem/cpu environment fails2022-04-13T23:08:02ZMatt devxz to produce base.tar.xz fails in low mem/cpu environment failsHi all,
After some troubleshooting where the build would fail for me (see https://bbs.archlinux.org/viewtopic.php?pid=2024067#p2024067), I found that xz is failing to produce the base.tar.xz file. Removing -T0 appears to fix the issue, ...Hi all,
After some troubleshooting where the build would fail for me (see https://bbs.archlinux.org/viewtopic.php?pid=2024067#p2024067), I found that xz is failing to produce the base.tar.xz file. Removing -T0 appears to fix the issue, though it can take quiet a while for the compression to complete, leaving -T0 and removing -9 also appears to fix the issue.
I'm attempting to build this container in a VM with 4 cores & 2 GB memory, though the amount of memory that's available to the build process would likely be far less as i'm running other applications as well.
Using zstd might be a good option instead of using xz here. Perhapes adding -v to the xz command may help show the progress at the very least if it's taking a while to complete.https://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/62What a nightmare trying to login here...2022-01-08T12:44:09ZAidan MPWhat a nightmare trying to login here...I came here to file a bug about LXC and then realized it's the wrong repo... sweet jesus tho, friends... if you're going to make me use an OTP then just do that and tie my authenticator to a username. Don't make me jump through endless h...I came here to file a bug about LXC and then realized it's the wrong repo... sweet jesus tho, friends... if you're going to make me use an OTP then just do that and tie my authenticator to a username. Don't make me jump through endless hoops with outdated technology when you're just going to make me use new technology anyway. Have a good one.https://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/61image contains too many packages?2021-12-13T22:12:08ZDieter Plaetinckimage contains too many packages?I'm probably missing something obvious, but the README says:
> images are tagged with the included package group / meta package and the timestamp of the date they were generated
Then when I compare the packages installed (+diffutils) w...I'm probably missing something obvious, but the README says:
> images are tagged with the included package group / meta package and the timestamp of the date they were generated
Then when I compare the packages installed (+diffutils) with what's in base, I get way more packages.
```
docker run -it library/archlinux:base
[root@d62b332e2894 /]# pacman -Sy diffutils && diff <(pacman -Qi base | sed -n 's/Depends On : //p' | tr ' ' '\n' | grep -v '^$' | sort) <(pacman -Q | sort | awk '{print $1}')
(...)
0a1,6
> acl
> archlinux-keyring
> argon2
> attr
> audit
> base
1a8
> brotli
2a10,12
> ca-certificates
> ca-certificates-mozilla
> ca-certificates-utils
3a14,20
> cryptsetup
> curl
> dbus
> device-mapper
> diffutils
> e2fsprogs
> expat
9a27
> glib2
10a29,32
> gmp
> gnupg
> gnutls
> gpgme
12a35,37
> hwids
> iana-etc
> icu
13a39
> iptables
14a41,76
> json-c
> kbd
> keyutils
> kmod
> krb5
> less
> libarchive
> libassuan
> libcap
> libcap-ng
> libcroco
> libelf
> libffi
> libgcrypt
> libgpg-error
> libidn2
> libksba
> libldap
> libmnl
> libnetfilter_conntrack
> libnfnetlink
> libnftnl
> libnghttp2
> libnl
> libp11-kit
> libpcap
> libpsl
> libsasl
> libseccomp
> libsecret
> libssh2
> libtasn1
> libtirpc
> libunistring
> libxcrypt
> libxml2
15a78,85
> linux-api-headers
> lz4
> mpfr
> ncurses
> nettle
> npth
> openssl
> p11-kit
16a87,89
> pacman-mirrorlist
> pam
> pambase
17a91,94
> pcre2
> pcre
> pinentry
> popt
19a97
> readline
21a100
> sqlite
22a102
> systemd-libs
24a105
> tzdata
25a107
> util-linux-libs
26a109,110
> zlib
> zstd
[root@d62b332e2894 /]#
```https://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/60/etc/hosts and /etc/resolv.conf should be added to NoExtract list2023-06-16T11:18:55ZClar Fon/etc/hosts and /etc/resolv.conf should be added to NoExtract listIf the `filesystem` package ever needs to be updated, these files will prevent the update from working since Docker mounts these files read-only.
EDIT: I've been testing this a bit, and it actually looks like there's a pacman bug, since...If the `filesystem` package ever needs to be updated, these files will prevent the update from working since Docker mounts these files read-only.
EDIT: I've been testing this a bit, and it actually looks like there's a pacman bug, since it's still failing to update even if I add the `NoExtract` options myself. The `CheckSpace` option doesn't actually honour the `NoExtract` options, meaning that if there is a read-only filesystem underneath those files, it will still fail. (This isn't default in this docker image, but it is on my config.)
Link to the pacman issue here: https://bugs.archlinux.org/task/72980https://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/59[Question] Has ArchLinux on Docker discontinued support for non-Indo-European...2022-08-22T14:17:13ZAlanIbr2049 shimz[Question] Has ArchLinux on Docker discontinued support for non-Indo-European locales?I recreated a docker Blackarch container today and found that I could not input Japanese characters correctly even if I started it with `-e TZ=Asia/Tokyo -e LANG=C.UTF-8 -e LANGUAGE=en_US:`. So I added `ja_JP.UTF-8 UTF-8` to /etc/locale....I recreated a docker Blackarch container today and found that I could not input Japanese characters correctly even if I started it with `-e TZ=Asia/Tokyo -e LANG=C.UTF-8 -e LANGUAGE=en_US:`. So I added `ja_JP.UTF-8 UTF-8` to /etc/locale.gen and tried to compile with the locale-gen command, but it was impossible.
```
[root@e4535bacdd21 /]# vim /etc/locale.gen
[root@e4535bacdd21 /]# locale-gen
Generating locales...
en_US.UTF-8... done
ja_JP.UTF-8... done [error] cannot open locale definition file `ja_JP': No such file or directory
```
Then I realized the problem in the first place. From the beginning, `/usr/share/i18n/locales/ja_JP` itself did not exist.
```
[root@e4535bacdd21 ~]# file /usr/share/i18n/locales/ja_JP
/usr/share/i18n/locales/ja_JP: cannot open `/usr/share/i18n/locales/ja_JP' (No such file or directory)
[root@e4535bacdd21 ~]# ls /usr/share/i18n/locales/
en_AG en_HK en_PH en_ZW translit_circle translit_fraction
en_AU en_IE en_SC i18n translit_cjk_compat translit_hangul
en_BW en_IL en_SG i18n_ctype translit_cjk_variants translit_narrow
en_CA en_IN en_US iso14651_t1 translit_combining translit_neutral
en_DK en_NG en_ZA iso14651_t1_common translit_compat translit_small
en_GB en_NZ en_ZM iso14651_t1_pinyin translit_font translit_wide
[root@e4535bacdd21 ~]#
```
I am studying information engineering at a college in Japan. The Blackarch Docker image is directly influenced by the ArchLinux Docker image, so I had to create a Blackarch container based on it for the report process. After some research, I think this originates from the ArchLinux side of the system, not from BlackArch.
I'll ask again. Is this a specification? Or is it a bug?https://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/58Keep the commented ParallelDownloads in pacman.conf2021-10-21T06:15:25ZAxel NavarroKeep the commented ParallelDownloads in pacman.confWhen the `/etc/pacman.conf` is edited for the docker with this content:
```
# We cannot check disk space from within a chroot environment
#CheckSpace
```
The parameter `#ParallelDownloads = 5` is fully removed from the file, but that c...When the `/etc/pacman.conf` is edited for the docker with this content:
```
# We cannot check disk space from within a chroot environment
#CheckSpace
```
The parameter `#ParallelDownloads = 5` is fully removed from the file, but that config line is in the `pacman.conf` at the root of this project.
Could we recheck that replace? I didn't find the code who is making this replace yet.https://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/57Cleanup package registry2021-04-17T14:19:28ZKristian KlausenCleanup package registryWe are building the images every day and pushing the rootfses to the [package registry](https://gitlab.archlinux.org/archlinux/archlinux-docker/-/packages) (~200MiB per day ~71GiB per year).
We only need to keep the rootfses used by the...We are building the images every day and pushing the rootfses to the [package registry](https://gitlab.archlinux.org/archlinux/archlinux-docker/-/packages) (~200MiB per day ~71GiB per year).
We only need to keep the rootfses used by the weekly releases around (~one year IIRC) and the daily rootfses can be deleted after a few days, let's say a month.https://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/56Unsupported host syscalls cause "failed to initialize alpm library"2023-08-03T11:19:36ZCaleb Maclennancaleb@alerque.comUnsupported host syscalls cause "failed to initialize alpm library"I regularly build Docker images with specialty tooling to power remote CI jobs using Arch Linux as a base. As of the last couple days all builds on Docker Hub have started to fail. I *was* using a two week old base image tag but I just t...I regularly build Docker images with specialty tooling to power remote CI jobs using Arch Linux as a base. As of the last couple days all builds on Docker Hub have started to fail. I *was* using a two week old base image tag but I just tried with the latest one and have the same problem. Local builds of the same Docker containers work fine.
I'm guessing (but not certain at this point) that systemd updates not liking Docker Hub's host environment have something to do with it again.
Here is the tail end of a build log showing the errors:
```
Step 7/26 : RUN pacman --needed --noconfirm -Syuq && yes | pacman -Sccq
---> Running in 20f78c0afb39
:: Synchronizing package databases...
downloading core.db...
downloading extra.db...
downloading community.db...
downloading alerque.db...
downloading alerque.db.sig...
:: Starting full system upgrade...
resolving dependencies...
looking for conflicting packages...
Packages (12) curl-7.75.0-1 e2fsprogs-1.46.0-1 findutils-4.8.0-1 gcc-libs-10.2.0-6 glib2-2.66.6-1 glibc-2.33-3 libldap-2.4.57-1 linux-api-headers-5.10.13-1 pacman-mirrorlist-20210206-1 systemd-247.3-1 systemd-libs-247.3-1 systemd-sysvcompat-247.3-1
Total Download Size: 54.81 MiB
Total Installed Size: 249.56 MiB
Net Upgrade Size: 0.04 MiB
:: Proceed with installation? [Y/n]
:: Retrieving packages...
downloading linux-api-headers-5.10.13-1-any.pkg.tar.zst...
downloading glibc-2.33-3-x86_64.pkg.tar.zst...
downloading gcc-libs-10.2.0-6-x86_64.pkg.tar.zst...
downloading findutils-4.8.0-1-x86_64.pkg.tar.zst...
downloading e2fsprogs-1.46.0-1-x86_64.pkg.tar.zst...
downloading libldap-2.4.57-1-x86_64.pkg.tar.zst...
downloading systemd-libs-247.3-1-x86_64.pkg.tar.zst...
downloading curl-7.75.0-1-x86_64.pkg.tar.zst...
downloading glib2-2.66.6-1-x86_64.pkg.tar.zst...
downloading pacman-mirrorlist-20210206-1-any.pkg.tar.zst...
downloading systemd-247.3-1-x86_64.pkg.tar.zst...
downloading systemd-sysvcompat-247.3-1-x86_64.pkg.tar.zst...
checking keyring...
checking package integrity...
loading package files...
checking for file conflicts...
:: Processing package changes...
upgrading linux-api-headers...
upgrading glibc...
Generating locales...
en_US.UTF-8... done
Generation complete.
upgrading gcc-libs...
upgrading findutils...
upgrading e2fsprogs...
upgrading libldap...
upgrading systemd-libs...
upgrading curl...
upgrading glib2...
upgrading pacman-mirrorlist...
�[91mwarning: /etc/pacman.d/mirrorlist installed as /etc/pacman.d/mirrorlist.pacnew
�[0mupgrading systemd...
upgrading systemd-sysvcompat...
:: Running post-transaction hooks...
(1/9) Creating system user accounts...
(2/9) Updating journal message catalog...
(3/9) Reloading system manager configuration...
Skipped: Current root is not booted.
(4/9) Updating udev hardware database...
(5/9) Applying kernel sysctl settings...
Skipped: Current root is not booted.
(6/9) Creating temporary files...
/usr/lib/tmpfiles.d/journal-nocow.conf:26: Failed to resolve specifier: uninitialized /etc detected, skipping
All rules containing unresolvable specifiers will be skipped.
(7/9) Reloading device manager configuration...
Skipped: Device manager is not running.
(8/9) Arming ConditionNeedsUpdate...
(9/9) Reloading system bus configuration...
Skipped: Current root is not booted.
�[91merror: failed to initialize alpm library
(could not find or read directory: /var/lib/pacman/)
�[0mThe command '/bin/sh -c pacman --needed --noconfirm -Syuq && yes | pacman -Sccq' returned a non-zero code: 255
build hook failed! (255)
```https://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/55Suggestion: restore excluded files in base image and provide a slim variant2023-06-23T00:25:01ZCédric ConnesSuggestion: restore excluded files in base image and provide a slim variantSince MR !13, pacman is configured to exclude non-essential files (such as docs and translations) in order to reduce the image size. At that time, a comment already mentioned that this makes the image less usable as a development environ...Since MR !13, pacman is configured to exclude non-essential files (such as docs and translations) in order to reduce the image size. At that time, a comment already mentioned that this makes the image less usable as a development environment. In addition, it seems to me to contradict one of the principles of the image:
> All installed packages have to be kept unmodified
Now that the image is listed as official, I would suggest to make it as close as possible to a regular Arch Linux installation. The current stripped version could become a slim variant, like the slim variant of the Debian image.
Would this make sense to you?Christian RebischkeSantiago Torres-Ariashashworkshashworks@archlinux.orgChristian Rebischkehttps://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/54DockerHub README shows ~ as strikeout2020-12-11T15:01:02Zhashworkshashworks@archlinux.orgDockerHub README shows ~ as strikeout![example](https://fb.hash.works/9f5Ha/)
Solution: Remove ~![example](https://fb.hash.works/9f5Ha/)
Solution: Remove ~hashworkshashworks@archlinux.orghashworkshashworks@archlinux.orghttps://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/53DockerHub archlinux/base redirect link requires login2020-12-03T18:23:44Zhashworkshashworks@archlinux.orgDockerHub archlinux/base redirect link requires loginIt should be `https://hub.docker.com/r/archlinux/archlinux` instead of `https://hub.docker.com/repository/docker/archlinux/archlinux`-It should be `https://hub.docker.com/r/archlinux/archlinux` instead of `https://hub.docker.com/repository/docker/archlinux/archlinux`-Sven-Hendrik Haasesvenstaro@archlinux.orgSven-Hendrik Haasesvenstaro@archlinux.orghttps://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/52[GitHub Issue] Missing lots of users and groups2020-12-11T15:08:31Zhashworkshashworks@archlinux.org[GitHub Issue] Missing lots of users and groupshttps://github.com/archlinux/archlinux-docker/pull/39
Maybe we need to move the `sysusers` call back to the `Dockerfile`. Also, we should add a test for this case.https://github.com/archlinux/archlinux-docker/pull/39
Maybe we need to move the `sysusers` call back to the `Dockerfile`. Also, we should add a test for this case.hashworkshashworks@archlinux.orghashworkshashworks@archlinux.orghttps://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/51[GitHub Issue] Parametrize 'docker' command2022-04-13T14:47:26Zhashworkshashworks@archlinux.org[GitHub Issue] Parametrize 'docker' commandhttps://github.com/archlinux/archlinux-docker/pull/38
Additionally, this should be added to the README.https://github.com/archlinux/archlinux-docker/pull/38
Additionally, this should be added to the README.https://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/50We can't push to archlinux/official-images because the PAT doesn't have the w...2020-11-23T11:50:19Zhashworkshashworks@archlinux.orgWe can't push to archlinux/official-images because the PAT doesn't have the workflow scopeJob [#9733](https://gitlab.archlinux.org/archlinux/archlinux-docker/-/jobs/9733) failed for 86fb81c860e8340b86a390f5d208556e3a3637e5:
```
$ git push -u origin "$head"
To https://github.com/archlinux/official-images.git
! [remote rejecte...Job [#9733](https://gitlab.archlinux.org/archlinux/archlinux-docker/-/jobs/9733) failed for 86fb81c860e8340b86a390f5d208556e3a3637e5:
```
$ git push -u origin "$head"
To https://github.com/archlinux/official-images.git
! [remote rejected] release/20201122.0.9724 -> release/20201122.0.9724 (refusing to allow a Personal Access Token to create or update workflow `.github/workflows/munge-pr.yml` without `workflow` scope)
error: failed to push some refs to 'https://github.com/archlinux/official-images.git'
```
We can't push to archlinux/official-images because the PAT doesn't have the workflow scope (needed to change the included file `.github/workflows/munge-pr.yml`).Frederik SchwanFrederik Schwanhttps://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/49docker-library would like us to link to the release in a comment2020-11-18T18:33:23Zhashworkshashworks@archlinux.orgdocker-library would like us to link to the release in a commenthttps://github.com/docker-library/official-images/pull/9009#issuecomment-721414221
> https://gitlab.archlinux.org/archlinux/archlinux-docker/-/package_files/125/download is a bit opaque as a download URL -- what do you think about addin...https://github.com/docker-library/official-images/pull/9009#issuecomment-721414221
> https://gitlab.archlinux.org/archlinux/archlinux-docker/-/package_files/125/download is a bit opaque as a download URL -- what do you think about adding https://gitlab.archlinux.org/archlinux/archlinux-docker/-/releases/v20201102.0.8296 in a comment to help users correlate that download URL with the "release" / other artifacts it's related to?https://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/48Make README more user-centric2020-11-30T21:17:15ZSven-Hendrik Haasesvenstaro@archlinux.orgMake README more user-centricCurrently our README.md is rather developer-centric. I think it should be user-centric first with a big "how to use our images" section on top and a developer-centric section rather towards the end.Currently our README.md is rather developer-centric. I think it should be user-centric first with a big "how to use our images" section on top and a developer-centric section rather towards the end.hashworkshashworks@archlinux.orghashworkshashworks@archlinux.orghttps://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/47Programmatically set READMEs on Dockerhub repositories2020-11-01T23:22:57ZSven-Hendrik Haasesvenstaro@archlinux.orgProgrammatically set READMEs on Dockerhub repositoriesThere's a README [here](https://hub.docker.com/repository/docker/archlinux/archlinux) and we should somehow automatically maintain that via API or such.There's a README [here](https://hub.docker.com/repository/docker/archlinux/archlinux) and we should somehow automatically maintain that via API or such.Kristian KlausenKristian Klausenhttps://gitlab.archlinux.org/archlinux/archlinux-docker/-/issues/46Mark archlinux/base as deprecated and remove it at a later date2021-04-29T00:58:48Zhashworkshashworks@archlinux.orgMark archlinux/base as deprecated and remove it at a later dateWe need to add a deprecation notice to https://hub.docker.com/r/archlinux/base since we now push to `archlinux/archlinux`.
At a later date we need to remove it, when would be a good date?
Also, while we are at it, `archlinux/archlinux`...We need to add a deprecation notice to https://hub.docker.com/r/archlinux/base since we now push to `archlinux/archlinux`.
At a later date we need to remove it, when would be a good date?
Also, while we are at it, `archlinux/archlinux` could use a nice logo, desc and links to this repo.Sven-Hendrik Haasesvenstaro@archlinux.orgSven-Hendrik Haasesvenstaro@archlinux.org2021-04-30