twa kernel panic under heavy load

Lowell Gilbert freebsd-questions-local at be-well.ilk.org
Wed Oct 5 06:27:55 PDT 2005


Dan Rue <drue at therub.org> writes:

> Greetings,
> 
> I am running a 3ware 9500 SATA raid card in a 12x300GB raid 50
> configuration.  Too often, I am seeing reboots during high I/O (rsync)
> operations.  
> 
> drue at leopard:~$ uname -a
> FreeBSD leopard.claimlynx.com 5.4-SECURITY FreeBSD 5.4-SECURITY #0: Thu
> Jun 30 02:25:52 UTC 2005
> root at builder.daemonology.net:/usr/obj/usr/src/sys/GENERIC  i386
> 
> 
> This is logged in /var/log/messages just before the reboot: 
> Oct  1 23:15:49 leopard kernel: twa0: Missing expected status bit(s) 2000<MC_RDY>
> Oct  1 23:15:49 leopard kernel: twa0: SCSI cmd = 0x2a: ERROR: (0x03:0x0001)
> Oct  1 23:15:49 leopard last message repeated 239 times
> Oct  1 23:15:50 leopard kernel: a: ERROR: (0x03: 0x0001)
> Oct  1 23:15:50 leopard kernel: twa0: SCSI cmd = 0x2a: ERROR: (0x03:0x0001)
> 
> 
> Here is dmesg identifying the controller: 
> 3ware device driver for 9000 series storage controllers, version: 2.50.02.012
> twa0: <3ware 9000 series Storage Controller> port 0xb800-0xb8ff mem 0xfb800000-0xfbffffff,0xfc5ffc00-0xfc5ffcff irq 24 at device 2.0 on pci2
> twa0: 12 ports, Firmware FE9X 2.06.00.009, BIOS BE9X 2.03.01.051
> 
> 
> Is this somethign that has been worked on in 5-STABLE or 6?  Would
> a kernel dump be helpful?  Please aim me at the appropriate list or
> people that would know.

The driver was imported from 3ware's own code on their website, and a
new version was imported since the 5.4 release was branched.  Your
problem may or may not have been fixed, but debugging it on the old
code base is definitely not worth anyone's time.

Definitely upgrade and see what happens.

Good luck.
-- 
Lowell Gilbert, embedded/networking software engineer, Boston area
		http://be-well.ilk.org/~lowell/


More information about the freebsd-questions mailing list