Vishnu
9 Posts
Quote from Vishnu on June 8, 2023, 9:30 am
I have a petasan cluster .In that cluster journal and cache physical disks failed.I need to replace the disks with new one without losing data.
I have a petasan cluster .In that cluster journal and cache physical disks failed.I need to replace the disks with new one without losing data.
admin
2,930 Posts
Quote from admin on June 8, 2023, 1:24 pm
When a journal disk has a hardware failure, its linked OSDs will fail, this is because the journal stores write ahead log (wal) as well as metadata database. The same is also true for cache since it is a write cache. When you replace the journal drive, you will need to delete then re-add the OSD disks from the UI as new OSDs.
The data on the OSDs will be lost, but due to 3x replication, you will have 2 other replicas of the data that was lost on 2 other nodes and the self healing process will re-recreate the lost replica.
When a journal disk has a hardware failure, its linked OSDs will fail, this is because the journal stores write ahead log (wal) as well as metadata database. The same is also true for cache since it is a write cache. When you replace the journal drive, you will need to delete then re-add the OSD disks from the UI as new OSDs.
The data on the OSDs will be lost, but due to 3x replication, you will have 2 other replicas of the data that was lost on 2 other nodes and the self healing process will re-recreate the lost replica.
Vishnu
9 Posts
Quote from Vishnu on June 12, 2023, 10:02 am
i replaced my journal and cache with 450GB SSD Disks.after adding journal and cache i added one 14.55TB OSD with external journal and cache.but i coudnt add my second osd disk.showing the error cache disk has no available space
i replaced my journal and cache with 450GB SSD Disks.after adding journal and cache i added one 14.55TB OSD with external journal and cache.but i coudnt add my second osd disk.showing the error cache disk has no available space
