Announcement Banner
UNSOLVED

SGibbs

updated

19 years ago

S

SGibbs

2 Posts

0

925

May 1st, 2007 03:00

PowerPath RHEL4 dmesg buffer I/O error on sd* device

Hi,

I've recently installed PP 4.5.1 on RHEL4 kernel version 2.6.9-42.0.10.ELsmp and wanted to confirm whether the following errors reported by dmesg are expected or not. The errors displayed take the form:

Buffer I/O error on device sdi, logical block 0
end_request: I/O error, dev sdi, sector 512
Buffer I/O error on device sdi, logical block 64
Buffer I/O error on device sdi, logical block 65
Buffer I/O error on device sdi, logical block 66
Buffer I/O error on device sdi, logical block 67
Buffer I/O error on device sdi, logical block 68
Buffer I/O error on device sdi, logical block 69
Buffer I/O error on device sdi, logical block 70
Buffer I/O error on device sdi, logical block 71
end_request: I/O error, dev sdi, sector 512
Buffer I/O error on device sdi, logical block 64

The errors relate to a number of different sd devices.....sdi, sdj, sdr, sdq, etc., but powermt display dev=all doesn't report doesn't display these devices under the I/O paths.

I can see the /dev/emcpower devices and have partitioned them, laid out a filesystem, mounted them and can write data so they seem to be OK.

The only thing I can think is that the sd devices displaying errors are only being reserved by an SP and not presented to the OS and if an I/O path failover were to occur the LUNs on that SP would become active.

Is that correct to assume?

Thanks,

Simon
  • admin111111

    70 Posts

    249

    0

    Posted May 1st, 2007 04:00

    Pls see Primus emc157139

    In summary:
    The devices logging the I/O error are assigned to SP-B, but currently "owned" by SP-A. The CLARiiON array is an active-passive array so this is normal behavior when multiple paths are utilized.
    These messages logged at boot up may be ignored.
  • SGibbs

    2 Posts

    249

    0

    Posted May 1st, 2007 07:00

    Thanks....that makes sense now.