Further clarification of the sync_action check behavior.

Hi!

	I have been reading the "Clarifications about check/repair, i.e.
RAID SCRUBBING" thread, and there were some answers which were still
slightly unclear to me, and I'd like to get a bit more clarification.
This is all in reference to the /sys/block/md0/md/sync_action setting.


Neil Brown wrote:

'check' just reads everything and doesn't trigger any writes unless a
read error is detected, in which case the normally read-error handing
kicks in.  So it can be useful on a read-only array.

'repair' does that same but when it finds an inconsistency is corrects
it by writing something.
If any raid personality had not be taught to specifically understand
'check', then a 'check' run would effect a 'repair'.  I think 2.6.17
will have all personalities doing the right thing.

check doesn't keep a record of problems, just a count.  'repair' will
reprocess the whole array.

What exactly is the normal read-error handling behavior that is invoked
when 'check' detects a read error, especially in regards to raid level
5?

Also, why exactly is 'check' useful on a read-only array? Does this mean
on an array which has 'array_state' set to 'readonly'? In this case, I'm
guessing this is designed to be useful in repairing a dirty and degraded
array, correct?

Lastly, under which mounted conditions are 'check' and 'repair'
safe/useful? Can I run both on a read-write mounted partition? Should
they be read-only, or totally unmounted? I'm assuming since resyncs can
occur while the overlying fs is mounted read-write, that it is both safe
and useful to both 'check' and 'repair' while the fs is mounted
read-write as well, but I don't see this mentioned explicitly anywhere.

Please CC: me, as I'm not on the list.

Thank you very much!
TJ Harrell