Netraid volume gets corrupted with Linux RH 7.2

[Domenico Viggiani]

Hi,
I have a LH3000 server with SCSI Netraid volumes and Linux RedHat 7.2 installed.
Under high load, a volume gets corrupted and I can see many SCSI errors on console.
Rebooting and putting disk online from Netraid BIOS solves the problem for a time.
I think it's a driver problem (historically Megaraid driver was a critical piece of software).

Any suggestion?

Domenico Viggiani

[G. Vrijhoeven]

Hi,

check out:


http://forums.itrc.hp.com/cm/QuestionAnswer/1,,0x69e4ff77de2bd611abd50090277a778c,00.html

Gideon

[Fred Skrotzki]

We tried for over a week and could not get it to work under EXTREAM load. We called the HP Support and all they could say was "use our driver" and update the bios on EVERYTHING (server, netriad, they even at one point said we needed to update the bios on the harddrives). I pointed out that 2.18 is the latest and it had been out and stable for 9 months but still we got the same speech.

We even flashed a card back to the AMI/LSI bios and still had issues so we gave up.

So that you know here is what we did:
We took the module supplied by HP and loaded the driver and modified the boot image, per directions. Also took the source supplied on disk 2 and compiled it directly into the kernel versions 2.4.2-2, 2.4.9-31, 2.4.12, and the latest 2.4.18. All would blow under extream load. You know when it's happening because suddenly you'll notice scsi_eh_? (where ? is the scsi card #) in the process stack and climbing.

Now I mean EXTREAM load. We were trying to build a series of LARGE mailing list servers and these need to send 8-10 millions of messages a day and receive about 2-3 million a day.

See http://forums.itrc.hp.com/cm/QuestionAnswer/1,,0xef7b42308663d611abdb0090277a778c,00.html for some more details on our exploits.

But we gave up about a week ago because nobody could solve the issue. We went back to software raid for now and will look into getting RAID controllers directly from LSI as those are quoted as working just fine.

[Domenico Viggiani]

Yes, I still have the same problem.
Also after upgrading all BIOS and firmware.

[Domenico Viggiani]

Did you get any success flashing card with AMI/LSI Megaraid 1600 Elite BIOS?