I have a ~ 38 GB table (MyISAM) and am using myisamchk -r to recover
it. Its been running for several days now (Win2k).
When I go to check it, it seems to have slowed down (sometimes hangs
without any CPU usage) and if I press the Enter key several times and
just fiddle around on the screen (open task manager, etc.) it seems
to "wake-up" myisamchk and it begins to check about 1000 recs every 1
to 2 seconds and happily creeps along.
I know I should have used explitict buffer sizes as I have plenty of
memory; but the performance of myisamchk doesn't seem to make sense.
The volume is raided and the raid manager shows the disks to be
healthy.
Unfortunately the temporary file is on the same volume, but I still
would expect a rather linear performance footprint of myisamchk.
mysqld is stopped.
There's 512MB ram.
connecting through remote desktop
The 4 drives are EIDE drives for a total of 207.88 GB space.
I have a syslog trapper running and storing data in an Access
database. But this is not CPU intensive either.
Again, the real concern is the performance footpring of myisamchk. It
shouldn't be so volatile.
Any ideas, similar experiences?
TIA