5 ms·
DRBD and Gluster are not any more resilient to filesystem corruption than a RAID device is. In this kind of case you hope for either real-time replicated storag
by imbriaco 13y ago
DRBD and Gluster are not any more resilient to filesystem corruption than a RAID device is. In this kind of case you hope for either real-time replicated storage on a completely separate physical host or very recent backups.
- dvanduzer 13y agoWhat are DRBD and Gluster if not real-time replicated storage on completely separate physical hosts? Filesystem corruption without hardware failure is far rarer in my experience. Have you seen an instance that wasn't a proverbial user error?
- imbriaco 13y agoYes, I have in fact, in a DRBD configuration. The bug was esoteric, but it happened and was not the result of user error. DRBD and Gluster both allow faults in the VFS layer to propagate to all replicas.
- avifreedman 13y agoGluster should by design I think avoid replicating filesystem metadata corruption (but would replicate internal metadata issues in files on top of the filesystem) but DRBD won't... At high volumes I still regularly break Gluster but it'd probably be OK for lower bandwidth/ops use. Not sure what the HN disk usage pattern is though.
- username223 13y agoIIRC Glusrerfs was the thing that gave me multiple identically-named files in the same directory. Useless.
- nilsbunger 13y agoYou never ran reiserfs I see... Back in ~2004 I watched IT spend a whole day recovering our 60-person startup's main Linux NFS server, due to a software bug in the storage driver. Had to rebuild the whole system from backups.
- sitkack 13y agohttp://basho.com/riak/ http://basho.com/riak/