lvm on HP-UX

[kandasamy]

How to find out bad disk in lvm on HP-UX 11 version


Thanks in Advance

[Torsten.]

This will guide you:

http://docs.hp.com/en/5991-1236/When_Good_Disks_Go_Bad.pdf

[James R. Ferguson]

Hi:

# diskinfo /dev/rdsk/cXtYdZ

...should return a non-zero size.

# dd if=/dev/rdsk/cXtYdZ of=/dev/null bs=1024k

...should complete without errors and an equal number of records in and out.

Note the use of the RAW device with 'dd' to bypass the buffer cache.

Regards!

...JRF...

[James George_1]

Hi

When disk go bad, you will start getting errors in your syslog file. Also, you can check the ioscan to see if it reports any NO_HW .

# ioscan -fnC disk

Also , check # dmesg

The first posting will help you to replace a bad disk.

Rgds / James

[Shardha]

Dear Kandasamy,


Monitoring for lbolt Messages in syslog.log
Also watch for lbolt messages in syslog.log, which often indicate that a disk or interface card needs to be replaced.
# tail /var/adm/syslog/syslog.log
vmunix: SCSI: Request Timeout; Abort -- lbolt: 137056, dev: 1f070500
vmunix: SCSI: Request Timeout; Abort -- lbolt: 167456, dev: 1f070500
The above message repeats 37 times.
Understanding Hexadecimal Hardware Addresses in syslog.log
The hexadecimal code at the end of the syslog SCSI error messages (1f070500 in the example
above) indicates which device may have caused the error.
The instructions below explain how to decode the hexadecimal address in the error message
to a disk device address. Consider the sample 1f070500 address mentioned in the message
above.
1f: The first two hexadecimal digits identify the driver that generated the error. In this
case, hexadecimal 1f is equivalent to decimal 31. lsdev tells us that 31 is the block
major# for the sdisk SCSI driver.
07: The next two digits identify the SCSI card instance number. In this case, hexadecimal
07 is equivalent to decimal 07, which indicates that the error came from one of the
/dev/dsk/c7txdx devices.
0: The next digit identifies the device target address. This error apparently came from
device /dev/dsk/c7t0dx.
5: The next two digits identify the LUN number. This error came from
/dev/dsk/c7t0d5.
00: The last two numbers identify device-specific options for tape drives. For disks, the last
two digits should always be 00.

Shardha

[tkc]

you can also check ems events coming from the disk_em monitor in the file /var/opt/resmon/log/event.log