For those not following source-changes@, I have just re-enabled the RAID 5 
discipline for softraid(4).

During the last two hackathons in Dunedin, the RAID 5 implementation was 
largely rewritten. As far as I am aware, the last missing part was the lack 
of ability to resume a partial rebuild, which has been fixed - it now needs 
further testing and usage so that any remaining issues can be found.

Unfortunately I do not currently have access to my D2 disk array, which has 
limited the amount of testing that I can do - while the basics of creating 
and using a RAID 5 volume certainly help, the really useful testing also 
includes disk failures, rebuilds and partial rebuilds, along with data 
verification.

At this point I would probably trust it enough to put real data on it, however 
I cannot promise that no data eating bugs still exist - obviously real world 
usage should have real world backups (and probably far more frequently than 
normal).

If you do test and find problems, please report them either on this list or 
directly if you prefer. I would also be more than happy to hear of any 
successes (providing sufficient beating has been applied first!).
-- 

   "Stop assuming that systems are secure unless demonstrated insecure;
    start assuming that systems are insecure unless designed securely."
          - Bruce Schneier

Reply via email to