Page MenuHomePhabricator

contint1001 and contint2001 need a newer version of Docker installed
Closed, ResolvedPublic

Description

The docker package installed on our contint* hosts is terribly out of date relative to the rest of the CI infrastructure and needs a newer package installed from thirdparty/ci.

The migration of contint1001 and contint2001 to buster effectively downgraded its docker package, since we'd been using thirdparty/ci for stretch (version 19.03) but started using the distro package with buster (version 18.09). Our runners are now running bullseye and docker 20.10. Since contint servers are the only runners allowed to publish images, their docker packages need to be kept as up-to-date as possible with other runners.

Event Timeline

Change 758986 had a related patch set uploaded (by Dduvall; author: Dduvall):

[operations/puppet@production] aptrepo: add docker packages to thirdparty/ci for buster

https://gerrit.wikimedia.org/r/758986

Change 758987 had a related patch set uploaded (by Dduvall; author: Dduvall):

[operations/puppet@production] contint: Install docker 20.10 from thirdparty/ci on buster

https://gerrit.wikimedia.org/r/758987

Mentioned in SAL (#wikimedia-operations) [2022-02-14T18:44:43Z] <mutante> contint2001 - disabling puppet, try replacing docker version (docker-io -> docker-ce), contint1001 first which is currently NOT the active server - gerrit:758987 T300682

Change 758986 merged by Muehlenhoff:

[operations/puppet@production] aptrepo: add docker packages to thirdparty/ci for buster

https://gerrit.wikimedia.org/r/758986

Change 763184 had a related patch set uploaded (by Muehlenhoff; author: Muehlenhoff):

[operations/puppet@production] Add missing update config for thirdparty/docker-ci-buster

https://gerrit.wikimedia.org/r/763184

Change 763184 merged by Muehlenhoff:

[operations/puppet@production] Add missing update config for thirdparty/docker-ci-buster

https://gerrit.wikimedia.org/r/763184

Change 758987 merged by Dzahn:

[operations/puppet@production] contint: Install docker 20.10 from thirdparty/ci on buster

https://gerrit.wikimedia.org/r/758987

Mentioned in SAL (#wikimedia-operations) [2022-02-16T16:51:15Z] <mutante> contint2001 - temp disabled puppet (active CI server) - contint1001 - attempting to install newer docker version (gerrit:758987 T300682)

@Muehlenhoff for some reason, it seems the docker-ce package did not make it into the thirdparty/ci component after https://gerrit.wikimedia.org/r/c/operations/puppet/+/758986 was merged.

Specifically:

I'm not sure whether I did something wrong in the reprepro configuration or there was an issue in deployment. Let me know if there's something I can do to fix it. Thanks!

And many thanks to @Dzahn for troubleshooting following the puppet apply of https://gerrit.wikimedia.org/r/758987 this morning.

@Muehlenhoff for some reason, it seems the docker-ce package did not make it into the thirdparty/ci component after https://gerrit.wikimedia.org/r/c/operations/puppet/+/758986 was merged.

Specifically:

I'm not sure whether I did something wrong in the reprepro configuration or there was an issue in deployment. Let me know if there's something I can do to fix it. Thanks!

https://gerrit.wikimedia.org/r/c/operations/puppet/+/758986 only adds a config where to pull from, but the sync still needs to be run by the reprepro command, I assume that didn't happen before https://gerrit.wikimedia.org/r/758987 was merged. I can import the packages in a bit.

These have now been imported:

jmm@gin:~$ curl -s https://apt.wikimedia.org/wikimedia/dists/buster-wikimedia/thirdparty/ci/binary-amd64/Packages | grep -e ^Package -e ^Version
Package: containerd.io
Version: 1.4.12-1
Package: docker-ce
Version: 5:20.10.12~3-0~debian-buster
Package: docker-ce-cli
Version: 5:20.10.12~3-0~debian-buster
Package: jenkins
Version: 2.319.3

Change 768774 had a related patch set uploaded (by Dduvall; author: Dduvall):

[operations/puppet@production] Revert "Revert "contint: Install docker 20.10 from thirdparty/ci on buster""

https://gerrit.wikimedia.org/r/768774

Change 768774 merged by Dzahn:

[operations/puppet@production] Revert "Revert "contint: Install docker 20.10 from thirdparty/ci on buster""

https://gerrit.wikimedia.org/r/768774

Change 789668 had a related patch set uploaded (by Dduvall; author: Dduvall):

[operations/puppet@production] contint: Bump docker 20.10 version for thirdparty/ci on buster

https://gerrit.wikimedia.org/r/789668

Change 789668 had a related patch set uploaded (by Dduvall; author: Dduvall):

[operations/puppet@production] contint: Bump docker 20.10 version for thirdparty/ci on buster

https://gerrit.wikimedia.org/r/789668

Change 789668 merged by Dzahn:

[operations/puppet@production] contint: Bump docker 20.10 version for thirdparty/ci on buster

https://gerrit.wikimedia.org/r/789668

Mentioned in SAL (#wikimedia-operations) [2022-05-05T18:42:01Z] <mutante> contitn2001 - apt-get remove --purge docker.io after docker-ce was installed by puppet for T300682

Mentioned in SAL (#wikimedia-operations) [2022-05-05T18:51:31Z] <mutante> contitn1001 - apt-get remove --purge docker.io after docker-ce was installed by puppet for T300682 (different behaviour from contint2001 since it did not have /var/lib/docker)

root@contint1001:/srv/docker# docker version
Client: Docker Engine - Community
 Version:           20.10.12
 API version:       1.41
 Go version:        go1.16.12
 Git commit:        e91ed57
 Built:             Mon Dec 13 11:45:37 2021
 OS/Arch:           linux/amd64
 Context:           default
 Experimental:      true

Server: Docker Engine - Community
 Engine:
  Version:          20.10.12
  API version:      1.41 (minimum version 1.12)
  Go version:       go1.16.12
  Git commit:       459d0df
  Built:            Mon Dec 13 11:43:46 2021
  OS/Arch:          linux/amd64
  Experimental:     false
 containerd:
  Version:          1.4.12
  GitCommit:        7b11cfaabd73bb80907dd23182b9347b4245eb5d
 runc:
  Version:          1.0.2
  GitCommit:        v1.0.2-0-g52b36a2
 docker-init:
  Version:          0.19.0
  GitCommit:        de40ad0
Dzahn claimed this task.

docker-ce replaced docker.io on both contint* machines andis on 20.10.12 now.