> Are you running the readbalancing patch?
>
Nope. I'm running the stock 2.2.17 kernel with raid-2.2.17-A0,
ide.2.2.17.all.20000904, the auto generated patches from i2c-2.5.2 and
lm_sensors-2.5.2, and a few other minor patches that I don't believe
could possibly be related to the problem.
> GFP_KERNEL appears twice in raid1.c did you replace both with GFP_BUFFER?
>
Yes, I did. I replaced both instances assuming a deadlock could occur with
either call to kmalloc(). The July 18th thread (and patch) which I based
this on modified both instances as well.
> I also swap on raid1 (64mb of physicial memory) so thought my problems might be
> related to that but will try your suggestion and see if it solves the problem
> are you swapping to raid1?
>
I do not swap on raid1.
My system consists of 4 IDE drives; each on seperate channels. The first
two are connected to the motherboard's controller while the second two
connected to a PCI IDE controller card (HPT366 chipset - hence the ide patch
above). Drive 0 (hda) contains a boot, root, and swap partition - neither
of which are under raid influence. Drives 1, 2, and 3 (hdc, hde, hdg
respectively) each contain a single partition which I use for my development.
I toggle their configuration from JBOD to RAID-0 to RAID-1 to RAID-5 more
than 20 times daily.
I'm still looking for an answer, which I will share with the raid community
when I find it.
-kral
============================================================================
Brian M. Kral [EMAIL PROTECTED]
ApplianceWare, Inc. www.applianceware.com
Fremont, CA
-
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to [EMAIL PROTECTED]
