Performance issues due to maintenance Saturday 2nd April 2022 15:36:00

We have temporarily added a new object store daemon to ph24-8 and managed to get the crashed ph24-8/9 object store daemon working again. The ph24-5 hypervisor was successfully taken out of service.

Migration of data that is still in progress may affect the performance of the Ceph cluster.

The ph24-8/9 object store daemon crashed while we were performing a data migration to other object stores. This failure resulted in unavailability of 25 pg's. The data was not lost, but was unavailable for a few minutes.