Page MenuHomePhabricator

Puppet failure on deployment-memc08.deployment-prep.eqiad.wmflabs
Closed, ResolvedPublic

Description

deployment-memc08 fails puppet. Apparently the instance got created a year ago as part of T213089: Upgrade memcached cluster to Debian Buster

The prompt shows:

The last Puppet run was at Fri Oct 30 18:00:44 UTC 2020 (9535 minutes ago).
/var/log/puppet.log
Using configured environment 'production'
Retrieving pluginfacts
Retrieving plugin
Retrieving locales
Loading facts
Could not retrieve catalog from remote server: Error 500 on SERVER: Server Error: Evaluation Error: Error while evaluating a Resource Statement, Evaluation Error: Error while evaluating a Function Call, Class[Profile::Memcached::Instance]:
  parameter 'size' expects an Integer value, got String
  parameter 'growth_factor' expects a Float value, got String (file: /etc/puppet/modules/role/manifests/memcached.pp, line: 9, column: 5) on node deployment-memc08.deployment-prep.eqiad.wmflabs
Not using cache on failed catalog
Could not retrieve catalog; skipping run

Event Timeline

hashar claimed this task.

Hiera at https://horizon.wikimedia.org/project/instances/70f350ff-1c96-45c0-8174-b13fd66a35c3/ has:

profile::memcached::extended_options:
- modern
profile::memcached::growth_factor: '1.15'
profile::memcached::size: '3000'
profile::memcached::version: present
puppetmaster: deployment-puppetmaster04.deployment-prep.eqiad.wmflabs

That hasn't been changed recently https://gerrit.wikimedia.org/g/cloud/instance-puppet/+log/master/deployment-prep/deployment-memc08.deployment-prep.eqiad1.wikimedia.cloud.yaml . I guess the Puppet class now enforces type.

I have dropped the single quotes surrounding the values (diff), and surely Puppet is now working.