Page MenuHomePhabricator

"publish-to-doc" jobs failing due to missing rsync on integration-agent-docker nodes
Closed, DuplicatePublicBUG REPORT

Description

Many failures listed at https://integration.wikimedia.org/ci/job/publish-to-doc/ starting at 2022-01-26T19:37.

00:00:00.153 + rsync --archive --stats --compress '--rsh=/usr/bin/ssh -a -T -o ConnectTimeout=6 -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no' jenkins-deploy@172.16.2.17:/srv/jenkins/workspace/workspace/mwext-node12-docs-docker-publish/src/docs/. .
00:00:00.361 Warning: Permanently added '172.16.2.17' (ECDSA) to the list of known hosts.
00:00:01.169 bash: line 1: rsync: command not found
00:00:01.169 rsync: connection unexpectedly closed (0 bytes received so far)

Event Timeline

This might be caused by the Bullseye base image being more bare bones than prior Debian base images. I don't see any obvious attempts to install rsync in the role::ci::slave::labs::docker Puppet module.

Sorry I have missed this task this morning. While reading the IRC backscroll I have noticed mediawiki-core-doxygen-docker failing and went to fill another task and fixed it T300236

My theory is that our Stretch instances had git-fat installed which depends on rsync. git-fat has later been removed but rsync was left behind. The Puppet manifests no more include git-fat and the fresh instances would not magically carry rsync > file not found. I have fixed it by ensuring rsync is present.