11-05-2007
Continueing relyability problems with SCO 6.0 on Dell Poweredge 1800 with PERC DC4.
Hello all,
I'am experiencing weird relyability problems with a SCO 6.0 Openserver server that runs on a Dell Poweredge 1800 equipped with a PERC DC4 raid controller and 4 36 GB 15 K rpm hot swappable scsi harddisks.
It runs a RAID 10 configuration.
It uses the mega hba, Revision 8.03a Release Date: 03 May 2005 driver for the PERC DC4 raid controller.
The machine runs as a breeze, very fast access to the filesystem but the problem is that it can run without a problem for months, e.g. 3 months but suddenly freezes then. I have to say that the problem becomes worse after the first freeze after a fresh installation.
I've had a script on it that every 10 seconds put the time + the output of a w and a ps -ef to the harddisk.
The script has created the last file before the system froze, but not written the content to it.
The file 10 seconds earlyer is complete.
Due to that observation it seems to me that at the moment the system freezes the filesystem (Raid?) stops.
I've been thinking to return to a single SCSI disk with this system or to go to another SCO6.0 certified RAID controller.
We've tested the RAID controller extensively in a windows environment, and the diagnostic tools that come with the Dell system don't find any errors in this system during tests.
I think that its due to the SCO driver that we've the problems, but I don't know for sure. Any help on this one?
W. Kind Regards,
Frederik
8 More Discussions You Might Find Interesting
1. Linux Benchmarks
Notes:
System Configuration: Dell Computer Corporation PowerEdge 4600 (4 X Intel(R) XEON(TM) CPU 2.00GHz
1988.782 MHz) hyperthreaded
System clock frequency: 99.0827 MHz
Memory size (approximate): 2559 Megabytes
========================= CPUs ========================
CPU # CPU... (0 Replies)
Discussion started by: tnorth
0 Replies
2. Linux Benchmarks
CPU/Speed: Dual PIII 533
Ram: 256Mb PC133
Motherboard: Unknown, Intel Chipset
Cache: 256k on board
Controller: PARC Raid 2/Si, AIC-7880U
Disk: 30Gb RAID 5
Load: 1 user. Clean boot, init 5 but X-Windows not loaded.
Kernel: See Benchmarks.
pgms: gcc 3.2.3
... (2 Replies)
Discussion started by: Garp
2 Replies
3. Solaris
hi, i am trying to install solaris 10 (on 4 cd's downloaded from the sun website) on my dell poweredge 2800 server. 2 xeon processors, 2GB of RAM and 2 NICS, 73 GB SCSI (maxtor ultra320, 3.5 series), i can give more details on request...
the installation crashes when trying to detect the NICS... (0 Replies)
Discussion started by: xinugeek
0 Replies
4. Red Hat
I have successfully installed RedHat 8.0 onto a Dell PowerEdge SC1425 today. This server has two SATA hard drives, and an IDE DVD-ROM drive.
Using the following kernel parameters, i successfully installed across both hard drives from CD:
ide0=0x1f0,0x3f6,14 vga=791 resolution=1024x768 expert... (5 Replies)
Discussion started by: fishsponge
5 Replies
5. Filesystems, Disks and Memory
Does anybody no how to get around the fact that the Red Hat installation does not see the SCSI hard drives. It lets you manually pick the proper device, but still fails saying no drive detected (or something to that effect). I was going to install windows 2003 Server and use VMware...but again,... (1 Reply)
Discussion started by: Fatflea
1 Replies
6. UNIX for Dummies Questions & Answers
I have 2 Dell Poweredge 1950 servers running . I have been having intermittent performance issues with the NIC cards on one of them. The two servers are identical and are running the same operating system. The server that has the issue is on the DMZ on a a static IP and is hosting a website. ... (0 Replies)
Discussion started by: skotapal
0 Replies
7. Programming
Hi,
I have a C++ program to access /dev/mem and retrieve details like Vendor, Manufacturer details of the motherboard. This works fine on all the machines except for on Dell Poweredge 2850,1950... machines.
I receive a 'EFAULT' when I try to access /dev/mem on these servers.
I suspect some... (1 Reply)
Discussion started by: ragisreekanth
1 Replies
8. SCO
I was asked to get a Dell PowerEdge 2600 server, out of warranty, running SCO 3.2v5.06 to perform a full backup. I spent hours researching and preparing a strategy only to fail on-site.
The only device names in /dev for a tape drive were
/dev/xStp0
/dev/xcdt0
/dev/xct0
... (6 Replies)
Discussion started by: ezlarry
6 Replies
AMR(4) BSD Kernel Interfaces Manual AMR(4)
NAME
amr -- AMI MegaRAID PCI-SCSI RAID driver
SYNOPSIS
amr* at pci? dev ? function ?
scsibus* at amr?
DESCRIPTION
The amr driver provides support for LSI (formerly American Megatrends) MegaRAID Express, Elite and Enterprise family RAID controllers for
SCSI and SATA, including models relabeled and sold by Dell, Hewlett-Packard, and Intel. Supported controllers include:
o MegaRAID 320-1
o MegaRAID 320-2
o MegaRAID Series 418
o MegaRAID Enterprise 1200 (Series 428)
o MegaRAID Enterprise 1300 (Series 434)
o MegaRAID Enterprise 1400 (Series 438)
o MegaRAID Enterprise 1500 (Series 467)
o MegaRAID Enterprise 1600 (Series 471)
o MegaRAID Elite 1500 (Series 467)
o MegaRAID Elite 1600 (Series 493)
o MegaRAID Express 100 (Series 466WS)
o MegaRAID Express 200 (Series 466)
o MegaRAID Express 300 (Series 490)
o LSI MegaRAID SCSI 320-0X, 320-2X, 320-4X
o LSI MegaRAID SCSI 320-1E, 320-2E
o LSI MegaRAID SATA 300-6x, 300-8x
o Dell PERC
o Dell PERC 2/SC
o Dell PERC 2/DC
o Dell PERC 4/Di
o Dell PERC 4/SC
o Dell PERC 4e/Si
o HP NetRAID-1/Si
o HP NetRAID-3/Si
o HP Embedded NetRAID
o Intel SRCU42X
o Intel SRCU42E
o Intel SRMOBU42E
o Intel SRCS28X
DIAGNOSTICS
Driver initialisation/shutdown phase
amr%d: memory window not available
amr%d: I/O window not available
The PCI BIOS did not allocate resources necessary for the correct operation of the controller. The driver cannot attach to this controller.
amr%d: busmaster bit not set, enabling
The PCI BIOS did not enable busmaster DMA, which is required for the correct operation of the controller. The driver has enabled this bit
and initialisation will proceed.
amr%d: can't allocate register window
amr%d: can't allocate interrupt
amr%d: can't set up interrupt
amr%d: can't allocate parent DMA tag
amr%d: can't allocate buffer DMA tag
amr%d: can't allocate scatter/gather DMA tag
amr%d: can't allocate s/g table
amr%d: can't allocate mailbox tag
amr%d: can't allocate mailbox memory
A resource allocation error occurred while initialising the driver; initialisation has failed and the driver will not attach to this con-
troller.
amr%d: can't obtain configuration data from controller
amr%d: can't obtain product data from controller
The driver was unable to obtain vital configuration data from the controller. Initialisation has failed and the driver will not attach to
this controller.
amr%d: can't establish configuration hook
amr%d: can't scan controller for drives
The scan for logical drives managed by the controller failed. No drives will be attached.
amr%d: device_add_child failed
amr%d: bus_generic_attach returned %d
Creation of the logical drive instances failed; attachment of one or more logical drives may have been aborted.
amr%d: flushing cache...
The controller cache is being flushed prior to shutdown or detach.
Operational diagnostics
amr%d: I/O beyond end of unit (%u,%d > %u)
A partitioning error or disk corruption has caused an I/O request beyond the end of the logical drive. This may also occur if FlexRAID Vir-
tual Sizing is enabled and an I/O operation is attempted on a portion of the virtual drive beyond the actual capacity available.
amr%d: polled command timeout
An initialisation command timed out. The initialisation process may fail as a result.
amr%d: bad slot %d completed
The controller reported completion of a command that the driver did not issue. This may result in data corruption, and suggests a hardware
or firmware problem with the system or controller.
amr%d: I/O error - %x
An I/O error has occurred.
SEE ALSO
cd(4), ch(4), intro(4), pci(4), scsi(4), sd(4), st(4), amrctl(8)
BSD
July 23, 2006 BSD