MSA2000 disks appear/disappear from RHEL

[Abdul Wahab]

We are having 2 DL 580 Servers with LINUX Ver 5 update 2 (latest) installed.
2 HBA's (HP SC44Ge, Part # 416096-B21) ON EACH SERVER CONNECTED TO EACH SAN SWITCH.
2 SAN SWITHES ARE INSTALLED - REDUNDANT CONFIGURATION.
MSA 2000 (2212FC DUAL enhanced CONTROLLER)

We have created 8 volumes and presented to both the servers(CLUSTER SET UP).
With out installing any drivers for the HBA, the external disks were detected. But there is a problem:

when we give the command # fdisk -l some times it shows the volumes presented and sometimes it hangs for few seconds and the volumes are not shown.
Again when we give the same command it shows the disk. if we continously give the command we can see the disk for 3 or 4 outputs then at the 5th output it hangs and we cannot view the volumes presented in the MSA.

We did a test..

we removed the fiber channel of one MSA controller and noticed that there was not issue and everything was ok.

But when we reconnnected the fiber cables back to the controller it the problem was back.

We tested both the controllers independently and both seems to be working fine independently.i.e it works fine when one raid controller path is available, it acts wierd when two controller paths are active.



we downloaded driver for SC44Ge for RHEL 5 x86 . when I tried installing it , it aborted saying the kernel is not in the supported list. (looks like our kernel is latest than what is defined in it's script).

sometimes its just like this:

[root@hancodb2 ~]# fdisk -l

Disk /dev/cciss/c0d0: 146.7 GB, 146778685440 bytes

255 heads, 63 sectors/track, 17844 cylinders

Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System

/dev/cciss/c0d0p1 * 1 13 104391 83 Linux

/dev/cciss/c0d0p2 14 3929 31455270 8e Linux LVM

/dev/cciss/c0d0p3 3930 6540 20972857+ 8e Linux LVM

/dev/cciss/c0d0p4 6541 17844 90799380 5 Extended

/dev/cciss/c0d0p5 6541 7845 10482381 8e Linux LVM

/dev/cciss/c0d0p6 7846 17844 80316936 8e Linux LVM

[root@hancodb2 ~]#

======

and after some seconds it appears like this:





[root@hancodb2 ~]# fdisk -l

Disk /dev/cciss/c0d0: 146.7 GB, 146778685440 bytes

255 heads, 63 sectors/track, 17844 cylinders

Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System

/dev/cciss/c0d0p1 * 1 13 104391 83 Linux

/dev/cciss/c0d0p2 14 3929 31455270 8e Linux LVM

/dev/cciss/c0d0p3 3930 6540 20972857+ 8e Linux LVM

/dev/cciss/c0d0p4 6541 17844 90799380 5 Extended

/dev/cciss/c0d0p5 6541 7845 10482381 8e Linux LVM

/dev/cciss/c0d0p6 7846 17844 80316936 8e Linux LVM

Disk /dev/sdi: 204.8 GB, 204800524288 bytes

255 heads, 63 sectors/track, 24898 cylinders

Units = cylinders of 16065 * 512 = 8225280 bytes

Disk /dev/sdi doesn't contain a valid partition table

Disk /dev/sdj: 51.2 GB, 51200917504 bytes

64 heads, 32 sectors/track, 48829 cylinders

Units = cylinders of 2048 * 512 = 1048576 bytes

Disk /dev/sdj doesn't contain a valid partition table

Disk /dev/sdk: 2048 MB, 2048917504 bytes

64 heads, 62 sectors/track, 1008 cylinders

Units = cylinders of 3968 * 512 = 2031616 bytes

Disk /dev/sdk doesn't contain a valid partition table

Disk /dev/sdl: 2048 MB, 2048917504 bytes

64 heads, 62 sectors/track, 1008 cylinders

Units = cylinders of 3968 * 512 = 2031616 bytes

Disk /dev/sdl doesn't contain a valid partition table

Disk /dev/sdm: 1024 MB, 1024458752 bytes

32 heads, 62 sectors/track, 1008 cylinders

Units = cylinders of 1984 * 512 = 1015808 bytes

Disk /dev/sdm doesn't contain a valid partition table

Disk /dev/sdn: 1024 MB, 1024458752 bytes

32 heads, 62 sectors/track, 1008 cylinders

Units = cylinders of 1984 * 512 = 1015808 bytes

Disk /dev/sdn doesn't contain a valid partition table

Disk /dev/sdo: 1024 MB, 1024458752 bytes

32 heads, 62 sectors/track, 1008 cylinders

Units = cylinders of 1984 * 512 = 1015808 bytes

Disk /dev/sdo doesn't contain a valid partition table

Disk /dev/sdp: 30.0 GB, 30000807936 bytes

64 heads, 32 sectors/track, 28611 cylinders

Units = cylinders of 2048 * 512 = 1048576 bytes

Disk /dev/sdp doesn't contain a valid partition table

[root@hancodb2 ~]#

[Phil uk · Accepted solution]

Hi,
There is a new release of firmware (29th Oct) which is considered "Critical" by HP.

It might be worth upgrading the firmware, if not already up-to-date, and then testing the MSA again.

http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareDescription.jsp?lang=en&cc=us&prodTypeId=12169&prodSeriesId=3687115&swItem=co-65944-1&prodNameId=3687125&swEnvOID=4004&swLang=13&taskId=135&mode=4&idx=0

Regards,
Phil

[Abdul Wahab]

Hello
Thanks for the reply.
We tried the firmware upgrade too before, but that didn't help.
Yesterday I solved the problem. The storage guy ,for some reason unknown, enabled 'interconnect' in the controllers. Interconnect should be enabled in case HBAs are directly connected to storage,But in our case we had it through SAN switches.
When I disabled interconnect and enabled 'straight through', everything was stable and fine.