topic Re: cciss messages in dmesg in Operating System - Linux

cciss messages in dmesg

Markus Schibli — Tue, 16 Jun 2009 13:27:23 GMT

Hi
could someone explain me the following cciss messages which I saw recently in dmesg?

Red Hat Enterprise Linux ES release 4 (Nahant Update 8)

dmesg output (partly):

HP CISS Driver (v 2.6.20.RH1)
cciss: Device 0x46 has been found at bus 4 dev 3 func 0
ACPI: PCI Interrupt 0000:04:03.0[A] -> GSI 51 (level, low) -> IRQ 217
cciss: using DAC cycles
blocks= 142253279 block_size= 512
heads= 255, sectors= 32, cylinders= 17433

blocks= 573480303 block_size= 512
heads= 255, sectors= 32, cylinders= 70280

blocks= 142253279 block_size= 512
heads= 255, sectors= 32, cylinders= 17433

cciss/c0d0: p1 p2
blocks= 573480303 block_size= 512
heads= 255, sectors= 32, cylinders= 70280

cciss/c0d1:
blocks= 573480303 block_size= 512
heads= 255, sectors= 32, cylinders= 70280

blocks= 573480303 block_size= 512
heads= 255, sectors= 32, cylinders= 70280

cciss/c0d1:
blocks= 573480303 block_size= 512
heads= 255, sectors= 32, cylinders= 70280

blocks= 573480303 block_size= 512
heads= 255, sectors= 32, cylinders= 70280

cciss/c0d1:
blocks= 573480303 block_size= 512
heads= 255, sectors= 32, cylinders= 70280

blocks= 573480303 block_size= 512
heads= 255, sectors= 32, cylinders= 70280

cciss/c0d1:
blocks= 573480303 block_size= 512
heads= 255, sectors= 32, cylinders= 70280

blocks= 573480303 block_size= 512
heads= 255, sectors= 32, cylinders= 70280

cciss/c0d1:
blocks= 573480303 block_size= 512
heads= 255, sectors= 32, cylinders= 70280

blocks= 573480303 block_size= 512
heads= 255, sectors= 32, cylinders= 70280
.
.
.
I can't find information what these messages mean. Any info on that? where to look after?

Many thanks in advance
Regards
Markus

Re: cciss messages in dmesg

Steven E. Protter — Tue, 16 Jun 2009 13:45:31 GMT

Shalom,

Disks have been detected, probably due to RH's built in multi-pathing capabilities.

Notice they are all the same size.

SEP

Re: cciss messages in dmesg

Markus Schibli — Tue, 16 Jun 2009 13:52:31 GMT

Hi Steven
Thanks for your reply.
I recognised too that the disks are all of the same size, in fact it's always the same disk.
It's always cciss/c0d1. Why? I saw ~180 exactly the same entries for cciss/c0d1 in dmesg.

Thanks and regards
Markus

Re: cciss messages in dmesg

John McNulty_2 — Tue, 16 Jun 2009 14:17:06 GMT

This can't be multipathing with cciss. It an internal RAID controller and only presents disks down one path. The messages are all related to the same device cciss/c0d1, and for the most part the same block number on that device. Maybe it's detected a bad block on the disk and is attempting to mapping it out and replace it with another?

I'm afraid I have no idea what these messages mean otherwise. Have you installed the Proliant Support Pack for your system. HP's diagnostic tools might throw more light on this for you.

Re: cciss messages in dmesg

macosta — Tue, 16 Jun 2009 22:20:13 GMT

I don't have the answer for this exact issue, but cciss can be for non-internal disks. For example, the cciss driver is used for the 700m SAS RAID controller which links to external arrays.

Re: cciss messages in dmesg

Markus Schibli — Fri, 26 Jun 2009 06:58:13 GMT

At last, I've not been able to exactly figure out why these messages appeared. It seemes that it's not possible in Linux to have the correct answer for each message which is logged by the os. Sometimes a little pain :-(