March 16, 201313 yr I might as well run a small badblocks nondestructive test once a month on each of my drives (one at a time), my question is:- Will writing to the drive also write to the parity with badblocks? The reason why I hope it won't is because that badblock deals on a physical level (Bits of sectors) while everything else that effects the parity works on a filesystem level (Bits of files), also the fact that:- A. It'd be nice to not slow the test down by a parity B. It'd be nice not to put 10 times the stress on my parity vs every other drive, asking for failures C. It'd be nice to have a backup of that drive if it were to get corrupted (E.G. rebuild the drive using the parity & rest of the disks) B. Parity isn't hardware based, where, I'm 100% sure it would effect it.
March 17, 201313 yr parity DOES NOT work on the filesystem level. It works on the bit level. Every bit on every sector in the first partition is involved in calculating parity. By running badblocks in non-destructive read-write mode, you ARE re-calculating parity for every write, and yes, parity will be written when any data disk is written.
March 17, 201313 yr Author parity DOES NOT work on the filesystem level. It works on the bit level. Every bit on every sector in the first partition is involved in calculating parity. By running badblocks in non-destructive read-write mode, you ARE re-calculating parity for every write, and yes, parity will be written when any data disk is written. Is there anyway to mark any drive attached to badblocks as "missing" until badblocks is done? If another drive fails, I can just cancel badblocks and recalculate from there, so, I don't believe it'll be too much of an issue? Correct me if I'm wrong. EDIT:- Just thought of what happens when it's done, it'll have to recalculate parity on the missing block so in fact if another drive were to fail, it'd be useless. I guess there's no way to do it without putting loads of pressure on my parity drive & slowing it down a load.
Archived
This topic is now archived and is closed to further replies.