On slave24 and slave25, /opt/qa has not been updated since May 2. Can we look into what's causing that?
I suspect the 2 servers were down during the last run of ansible-playbook, but this is also supposed to be run at night. I will dig the log.
So on slave24: fatal: [slave24.cloud.gluster.org]: FAILED! => {"changed": false, "cmd": "/usr/bin/git ls-remote origin -h refs/heads/master", "failed": true, "msg": "", "rc": 128, "stderr": "", "stdout": "", "stdout_lines": []} I test slave25 now.
So slave25 is the same, but not slave29. I didn't found yet any relevent difference. It also work when run from my laptop and nigel one (the ansible playbook, that's it). I strongly suspect a issue on the salt bus.
I actually noticed this on slave22 right now. There was no effect of git pull on the folder. I tried to rm -r /opt/qa and do a fresh git clone. The rm worked, the fresh clone didn't work. [root@slave22 opt]# git clone https://github.com/gluster/glusterfs-patch-acceptance-tests.git qa Cloning into 'qa'... [root@slave22 opt]# ls -l total 0
[root@slave22 qa]# GIT_TRACE=1 git pull trace: exec: 'git-pull' trace: run_command: 'git-pull' trace: built-in: git 'rev-parse' '--git-dir' trace: built-in: git 'rev-parse' '--is-bare-repository' trace: built-in: git 'rev-parse' '--show-toplevel' trace: built-in: git 'ls-files' '-u' trace: built-in: git 'symbolic-ref' '-q' 'HEAD' trace: built-in: git 'config' '--bool' 'branch.master.rebase' trace: built-in: git 'config' '--bool' 'pull.rebase' trace: built-in: git 'rev-parse' '-q' '--verify' 'HEAD' trace: built-in: git 'fetch' '--update-head-ok' trace: run_command: 'git-remote-https' 'origin' 'https://github.com/gluster/glusterfs-patch-acceptance-tests.git'
This just gets weirder and weirder. I can clone from gerrit successfully. It's just github that has an issue.
Ha, I just realized what's going on. We're running into this issue. Seemingly the fix didn't apply to this machine. https://access.redhat.com/solutions/2313911
Closing this now that we have the root cause.