1. Re-created disk

   Hide full report Show full report

   We've re-created one disk which kept causing hiccups.

   We do not expect any further noticeable performance impact.

2. Cause known

   Hide full report Show full report

   The amount of stored data has grown with around 1 TiB in the past 24 hours. This is causing us to hit Ceph bug 71235.

   We are adding more disks to decrease the amount of allocated blocks per disk.

3. Temporarily lower performance

   Hide full report Show full report

   Due to a bug in Ceph, the file system, multiple disks have crashed. We are currently replacing said disks, which causes temporarily lower performance while all data is copied.

   Replacing the disks will take several more hours, but each performance spike should take no longer than several minutes.

Experiencing an issue?