Page MenuHomePhabricator

cumin on deployment-cumin is broken
Closed, ResolvedPublic


krenair@deployment-cumin:~$ sudo cumin '*' run-puppet-agent
Traceback (most recent call last):
  File "/usr/bin/cumin", line 11, in <module>
    load_entry_point('cumin==2.0.0', 'console_scripts', 'cumin')()
  File "/usr/lib/python2.7/dist-packages/", line 356, in load_entry_point
    return get_distribution(dist).load_entry_point(group, name)
  File "/usr/lib/python2.7/dist-packages/", line 2476, in load_entry_point
    return ep.load()
  File "/usr/lib/python2.7/dist-packages/", line 2190, in load
  File "/usr/lib/python2.7/dist-packages/cumin/", line 16, in <module>
    from ClusterShell.NodeSet import NodeSet
ImportError: No module named ClusterShell.NodeSet

Event Timeline

I found this in /var/log/apt/history.log:

Start-Date: 2018-07-20  06:29:50
Commandline: /usr/bin/unattended-upgrade
Install: python3-clustershell:amd64 (1.8-1~wmf1, automatic)
Upgrade: clustershell:amd64 (1.7.3-1~wmf1, 1.8-1~wmf1)
End-Date: 2018-07-20  06:29:53

Ah, the cumin package was being kept back and was therefore out of date:

root@deployment-cumin:~# apt-cache policy cumin
  Installed: 2.0.0-1
  Candidate: 3.0.1-1
  Version table:
     3.0.1-1 0
       1001 jessie-wikimedia/main amd64 Packages
 *** 2.0.0-1 0
        100 /var/lib/dpkg/status

I've made it upgrade that and it functions again.

Krenair claimed this task.

Sorry I have forgot to take care of the deployment-prep cumin master :/ In T188112 volans did take care of the puppet boilerplate but I only looked at the integration cumin master. At least all puppet patches got merged, so that was just pending the package upgrade.

According to, cumin is also used by af-puppetmaster02.automation-framework.eqiad.wmflabs as well as (deployment|integration)-cumin but I assume @Volans took care of that.