Page MenuHomePhabricator

Create puppet profiles for the new ceph cluster
Closed, ResolvedPublic

Description

We have decided to build upon the existing module that we have in puppet for our new ceph cluster.

However, the profiles were had were too highly customized to the needs of the WMCS cluster, so these were renamed from profile::ceph to profile::cloudceph as these names more closely matched the machines to which they are applied.

This means that we can now make new ceph profiles for the new cluster. They should be made as generic as possible, so that ideally we can migrate the cloudceph and any other clusters to use them in future too.

Event Timeline

BTullis created this task.
BTullis added a subscriber: EChetty.

@EChetty this needs to be in the current sprint because it's the next logical piece of work on this project. Adding the tag, so I hope that's OK.

@EChetty I see that this is tagged as an Epic. Is this an Epic? If so, this one should be broken down because epics are almost always delivered over a set of sprints. If not, please, ignore the comment.

Change 887419 had a related patch set uploaded (by Btullis; author: Btullis):

[operations/puppet@production] Configure the new ceph monitor servers [WIP]

https://gerrit.wikimedia.org/r/887419

@JArguello-WMF - I've removed the epic tag. I think that was probably my fault. when I created it as a child ticket, it inherited the tags of the parent ticket including the epic tag.

Change 891854 had a related patch set uploaded (by Btullis; author: Btullis):

[labs/private@master] Add dummy keydata for the new ceph admin user

https://gerrit.wikimedia.org/r/891854

Change 891854 merged by Btullis:

[labs/private@master] Add dummy keydata for the new ceph admin user

https://gerrit.wikimedia.org/r/891854

Change 887419 merged by Btullis:

[operations/puppet@production] Configure the new ceph servers with mon and mgr daemons

https://gerrit.wikimedia.org/r/887419

Mentioned in SAL (#wikimedia-analytics) [2023-03-14T14:57:22Z] <btullis> deploying ceph mon and mgr daemons to cephosd100[1-5] T328123

Icinga downtime and Alertmanager silence (ID=d004b3da-f6b2-44bc-994d-9e4ff6dc6413) set by btullis@cumin1001 for 1 day, 12:00:00 on 5 host(s) and their services with reason: Bootstrapping ceph

cephosd[1001-1005].eqiad.wmnet