tech-userlevel archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: swap-on-raidframe vs raidctl -P

der Mouse writes:
> >> Swap is on raid0b, and, when raidframeparity ran, it complained that
> >> raid0 parity was dirty and started a parity rewrite.  [...]
> > Without touching on your other points below, do you have swapoff=YES
> > in /etc/rc.conf?
> rc.conf does not mention swapoff at all (/etc/defaults/rc.conf does
> contain "swapoff=YES", and no other references to swapoff).
> How is that relevant, though?  This is happening at startup, not
> shutdown.

Whether a parity check is required at startup is effectively 
determined at shutdown.  If the shutdown happens cleanly, then the 
check/rewrite will not be required on startup.

(Assume a RAID set is "clean".  On the first open of a partition on a 
RAID set, the component labels will be updated to say it its "DIRTY", 
even though the kernel knows things are clean.  On the last close of
the last open partition on a RAID set, the 'internal state' will be 
written out to the component labels.  In this example, that would be 
'clean'.  If swap does not get unconfigured correctly, then that 
'last open partition' never gets closed, and the final updating of 
the component labels doesn't happen.  That is what results in the
RAID set being marked 'DIRTY', and requiring the rewrite/check, even
theough the RAID set is very likely to be completely clean.)


Greg Oster

Home | Main Index | Thread Index | Old Index