Problem with 2.4.19-pre10-ac2 and i2o SX6000 controller

Problem with 2.4.19-pre10-ac2 and i2o SX6000 controller

Post by Aubrey Lewm » Tue, 18 Jun 2002 11:55:17



I've been unable to get my Promise SX6000 IDE raid controller card to
behave in my linux system for several months now and I don't know what
I should do to resolve the problem.

My system is a Tyan Thunder K7 - S2462 board with 512Mb Reg PC 2100
DDR and dual MP 1800+ cpus. (I've tried other cpu's (single) with the
same results).  It is the lower end model without built in scsi.  Its
bios is currently 2.10.  I'm using the built in nic and video card.
In fact, the only additional card I have is the Promise SX6000 IDE
RAID controller card.

The raid controller card has 5 Maxtor 80 gig drives installed with 4
in Raid 5 and one as a hot spare.

I'm running RedHat 7.2 as my base system and this is my unmame output:


Linux supertux 2.4.19-pre10-ac2 #2 SMP Mon Jun 10 09:12:59 PDT 2002
i686 unknown

Anyway, the biggest problem is that after some period of time (a day
or three), my disk array will go out to lunch and processes won't be
able to read/write files anymore.  The system doesn't hang as it is
still responsive until it has to process an i/o. (like login prompts
will accept a user to log in as, but hang when actually trying to log
in)

I'm also unable to soft reboot the system as once the i2o subsystem
initializes the controller, the controller will not be able to
successfully reinitialize the array when the system is reset.  What I
mean by that is that the SX6000 bios will not see a defined array
again until the system is powered off and restarted.  Totally lame,
yes.

Also, when copying files around the system, the whole OS will pause
for several seconds sometimes.. All processes on the system are just
frozen for a while, and then things go on like normal after 2-8
seconds.

My /var/log/messages file is filled with these:
Jun 16 11:54:55 supertux kernel: i2o/iop0: Hardware Failure: Unknown
Error
Jun 16 11:54:55 supertux kernel: i2o/iop0: Hardware Failure: Unknown
Error
Jun 16 11:55:28 supertux last message repeated 2 times
Jun 16 11:55:28 supertux kernel: i2o/iop0: Hardware Failure: Unknown
Error
Jun 16 11:55:28 supertux kernel: i2o/iop0: No handler for event
(0x00000020)
Jun 16 11:55:28 supertux last message repeated 6 times

When my system starts up, the i2o section looks like:
Jun 16 10:20:23 supertux kernel: I2O Core - (C) Copyright 1999 Red Hat
Software
Jun 16 10:20:23 supertux kernel: I2O: Event thread created as pid 14
Jun 16 10:20:23 supertux kernel: Linux I2O PCI support (c) 1999 Red
Hat Software.
Jun 16 10:20:23 supertux kernel: i2o: Checking for PCI I2O
controllers...
Jun 16 10:20:23 supertux kernel: i2o: I2O controller on bus 0 at 73.
Jun 16 10:20:23 supertux kernel: i2o: PCI I2O controller at 0xFC400000
size=4194304
Jun 16 10:20:23 supertux kernel: I2O: Promise workarounds activated.
Jun 16 10:20:23 supertux kernel: i2o/iop0: Installed at IRQ10
Jun 16 10:20:23 supertux kernel: i2o: 1 I2O controller found and
installed.
Jun 16 10:20:23 supertux kernel: Activating I2O controllers...
Jun 16 10:20:23 supertux kernel: This may take a few minutes if there
are many devices
Jun 16 10:20:23 supertux keytable: Loading system font:  succeeded
Jun 16 10:20:23 supertux kernel: i2o/iop0: Reset rejected, trying to
clear
Jun 16 10:20:23 supertux kernel: i2o/iop0: LCT has 9 entries.
Jun 16 10:20:23 supertux kernel: i2o/iop0: Configuration dialog
desired.
Jun 16 10:20:23 supertux kernel: I2O configuration manager v 0.04.
Jun 16 10:20:23 supertux kernel:   (C) Copyright 1999 Red Hat Software
Jun 16 10:20:23 supertux kernel: I2O Block Storage OSM v0.9
Jun 16 10:20:23 supertux kernel:    (c) Copyright 1999-2001 Red Hat
Software.
Jun 16 10:20:23 supertux kernel: i2o_block: Checking for Boot
device...
Jun 16 10:20:23 supertux kernel: i2o_block: Checking for I2O Block
devices...
Jun 16 10:20:23 supertux kernel: i2ob: Installing tid 14 device at
unit 0
Jun 16 10:20:23 supertux kernel: i2o/hda: Max segments 28, queue depth
8, byte limit 49152.
Jun 16 10:20:23 supertux kernel: i2o/hda: Type 62: 234501MB, 512 byte
sectors.
Jun 16 10:20:23 supertux kernel: i2o/hda: Maximum sectors/read set to
96.
Jun 16 10:20:23 supertux kernel: Partition check:
Jun 16 10:20:23 supertux kernel:  i2o/hda: i2o/hda1 i2o/hda2 i2o/hda3
i2o/hda4 < i2o/hda5 i2o/hda
6 i2o/hda7 i2o/hda8 i2o/hda9 i2o/hda10 i2o/hda11 >

I have no idea what the 'hardware failure' message is indicating, but
it doesn't give me warm fuzzy feelings. When the system hangs because
of i/o failure, there are no log entries indicating any other issues
besides the hardware failure or no handler for event messages that I
showed.

I've asked promise tech support about this controller and the response
I got back was it should work with this board, but that there are
known issues with the 760MPX based boards.

Anyway, does anyone have any idea of what I should do/try or what
might be going wrong?  Does anyone have any success stories with a
SX6000 on a 760MP chipset?

Thanks,

Aubrey Lewman