cluster

cluster

Resizing placement groups - again

Ceph really keeps us active. 😄

cluster

Upgrade Ceph from Quincy to Reef

We need to upgrade our ceph version from 17 (quincy) to 18 (reef) to be prepared for next major Proxmox upgrade.

cluster

Cleanup of ceph configs

After many manual changes our ceph configs still contains a few remainings that we now cleanup.

cluster

Shrink placement groups again

After the removal of nodes cluster-02 and cluster-06 the number of placement groups had to be optimized again.

cluster

Remove node cluster-02

It's time to remove the old cluster-02 node.

cluster

Add Ceph monitor and manager on node 12.

We add a Ceph monitor and a Ceph manager spare on node 12 using the web ui.

cluster

Shrink ssd pool placement groups

This time we have a 'too many placement groups' message on our ssd pool.

cluster-06

Remove cluster-06 node

We need to remove the cluster-06 node which was just a temporary replacement for the bricked cluster-01.

cluster

Node joins cluster 'minati'

We now the cluster with the new node.

cluster-12

Setup the network interfaces of node cluster-12

After installing the cluster-12 node we change the temporary network settings to the real ones.

cluster

Major ceph outage on cluster

Our cluster sufferes a major ceph breakdown.

cluster-12

Install Proxmox on cluster-12

Install Proxmox VE 8.1 on the cluster-12 hardware using the terminal installer.

cluster

Fix crashed osd

After the CEPH_CLUSTER_LAN outage we had a crashed OSD, which needs to be cleaned up.

cluster

Major crash

Or cluster has a major outage without any warning.

cluster

Remove Ceph hdd pool

As all vm disks have been migrated to the ssd pool we can now remove the hdd pool an it's OSDs.

cluster

Resize ssd pool placement groups once more

We currently have a 'too few placement groups' message on our ssd pool.