[4284] in linux-scsi channel archive
DPT 3334uw trouble
daemon@ATHENA.MIT.EDU (Jon Lewis)
Fri Jul 3 23:18:53 1998
Date: Fri, 3 Jul 1998 23:17:29 -0400 (EDT)
From: Jon Lewis <jlewis@inorganic5.fdt.net>
To: Linux SCSI Mailing List <linux-scsi@vger.rutgers.edu>
I've got a DPT 3334uw and a stack of IBM DDRS-34560 disks in a Granite
Digital rack enclosure that our NT person was having trouble with. To
help see if it was hardware or software (and since I need to build
something similar for a Linux system real soon), I tore it apart, verified
all jumper settings on the disks were ok, and stuck the DPT card in a P100
linux box.
I had no trouble creating a 13gb RAID-5 array for Linux using the DOS
dptmgr program, booted up linux, and loaded the eata_dma driver. The
eata_pio and eata drivers don't seem to like the 3334uw, and fail to load
if I try one of them instead. So, using the eata_dma driver, I partition
the array into 2 6.5gb ext2 partitions, made filesystems on them, mounted
them, and started running some disk tests. I use a program that spits out
random chars to STDOUT redirected to a file...and had no trouble creating
1mb, and 10mb files. When I try creating 100mb files, I get lots of eata
resets followed by an unexpected reboot.
This is fairly similar to what was happening when the NT guy was trying to
install NT. NT would fail to copy files and complain about disk errors,
or just spontaneously reboot.
We're wondering if we maybe have a bad card, bad RAM, or if maybe we
should just try another brand (ICP or Mylex). I'll try to setup the linux
box with remote syslogging so I can actually capture some of the reset
messages before the reboot. So far, nothing's gotten logged to disk.
------------------------------------------------------------------
Jon Lewis <jlewis@fdt.net> | Spammers will be winnuked or
Network Administrator | drawn and quartered...whichever
Florida Digital Turnpike | is more convenient.
______http://inorganic5.fdt.net/~jlewis/pgp for PGP public key____
-
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@vger.rutgers.edu