device-mapper: multipath: Failing path

[UNIX_HP]

hi all,
often i'm having problems with 6 server HP blade with linux 5.5,
in dmesg i found this error:

sd 0:0:0:0: SCSI error: return code = 0x00070000
end_request: I/O error, dev sds, sector 5582848
device-mapper: multipath: Failing path 65:32.
sd 0:0:0:0: SCSI error: return code = 0x00070000
end_request: I/O error, dev sds, sector 1717648
sd 0:0:0:0: SCSI error: return code = 0x00070000
end_request: I/O error, dev sds, sector 5587968
device-mapper: multipath: Failing path 65:32.
sd 0:0:0:0: SCSI error: return code = 0x00070000
end_request: I/O error, dev sds, sector 5588992
sd 0:0:0:0: SCSI error: return code = 0x00070000
end_request: I/O error, dev sds, sector 59049440
device-mapper: multipath: Failing path 65:32.
sd 0:0:0:0: SCSI error: return code = 0x00070000
end_request: I/O error, dev sds, sector 5598208
sd 0:0:0:0: SCSI error: return code = 0x00070000
end_request: I/O error, dev sds, sector 5600256
device-mapper: multipath: Failing path 65:32.
sd 0:0:0:0: SCSI error: return code = 0x00070000
end_request: I/O error, dev sds, sector 63225856
sd 0:0:0:0: SCSI error: return code = 0x00070000
end_request: I/O error, dev sds, sector 5607424
device-mapper: multipath: Failing path 65:32.
qla2xxx 0000:06:00.0: scsi(0:0:3): Abort command issued -- 1 a3be28 2002.
qla2xxx 0000:06:00.0: scsi(0:0:1): Abort command issued -- 1 a3be86 2002.
qla2xxx 0000:06:00.0: scsi(0:0:0): Abort command issued -- 1 a3bef2 2002.
qla2xxx 0000:06:00.0: scsi(0:0:1): Abort command issued -- 1 a3c0ca 2002.
qla2xxx 0000:06:00.0: scsi(0:0:5): Abort command issued -- 1 a3c26e 2002.
qla2xxx 0000:06:00.0: scsi(0:0:5): Abort command issued -- 1 a3c2d3 2002.
sd 0:0:0:0: SCSI error: return code = 0x00070000
end_request: I/O error, dev sds, sector 5607424
device-mapper: multipath: Failing path 65:32.


this error is the same in other 5 servers..
when i try to do a VGSCAN it freeze,

can anyone help me ?

Thanks

[Matti_Kurkela]

You seem to be using device-mapper-multipath (dm-multipath for short). The error messages seem to indicate some sort of trouble with your first QLogic HBA.

> sd 0:0:0:0: SCSI error: return code = 0x00070000

This code seems to mean "internal error", reported by the HBA driver. So the problem would most likely be in the driver, firmware or hardware of the HBA.

Go to the support website of your SAN manufacturer and find the compatibility list/database. It may list some specific minimum version requirements for the Linux kernel, HBA driver, HBA firmware or storage firmware. It may also list some requirements on HBA or storage configuration. Make sure all these requirements are met.

MK

[hari07]

https://access.redhat.com/knowledge/solutions/131553

[hari07]

environment effected 

• Red Hat Enterprise Linux 5 (RHEL5) update 3 and earlier
• QLogic HBA card using qla2xxx driver provided by Red Hat Enterprise Linux kernel.
• device-mapper-multipath to manage multiple SAN paths.

Resoln:
Upgrade to kernel-2.6.18-164.el5 which contains a fix, allowing specific errors to be retried without failing the path.
This issue is also fixed in rhel‑5.3.z kernel-2.6.18-128.1.10.el5.

[Gerardo_Arceri]

Please check the firmware on the HBA, we had similar problems and they were gone after upgrading the HBA firmware .