Page MenuHomePhabricator

Add 4 new Kubernetes worker nodes to ml-serve-eqiad
Closed, ResolvedPublic

Description

We have 4 new k8s worker nodes racked and configured: ml-serve100[5-8]

We should:

  1. Check partitions/os/etc.. on all nodes
  2. Add them to puppet
  3. Add them to homer (BGP configs)

Event Timeline

Change 784701 had a related patch set uploaded (by Elukey; author: Elukey):

[operations/puppet@production] Add four new k8s worker nodes to ml-serve-eqiad

https://gerrit.wikimedia.org/r/784701

Change 784703 had a related patch set uploaded (by Elukey; author: Elukey):

[operations/homer/public@master] Add ml-serve100[5-8] to the ml-serve-eqiad k8s BGP neighbors

https://gerrit.wikimedia.org/r/784703

Change 784701 merged by Elukey:

[operations/puppet@production] Add four new k8s worker nodes to ml-serve-eqiad

https://gerrit.wikimedia.org/r/784701

elukey changed the task status from Open to Stalled.Apr 26 2022, 7:23 AM

The new nodes are in row E/F, that have a different network configuration. Task blocked until T306649 is solved.

Change 784703 merged by jenkins-bot:

[operations/homer/public@master] Add ml-serve100[5-8] to the ml-serve-eqiad k8s BGP neighbors

https://gerrit.wikimedia.org/r/784703

elukey claimed this task.

The hosts are working, I am going to follow up in https://phabricator.wikimedia.org/T306649 to improve the network config.