dumpmdiskbadblocks
Use the dumpmdiskbadblocks command to write the bad block counts and locations that are on a specified MDisk to a dump file for use by fix procedures.
Parameters
- object_id | object_name
- (Required) Specifies the MDisk for which you need to dump the bad block record table.
Description
Use the dumpmdiskbadblocks command to write the bad block counts and locations that are on a specified MDisk to a readable ASCII dump file for use by fix procedures. The output consists of bad blocks for which an event log has been raised.
Use lsdumps -prefix /dumps/mdisk to list the output files. Use cleardumps -prefix /dumps/mdisk to clear the output files.
The reported event log sequence numbers correspond to the first event seen in the bad block
record, which is a 512-block region.
- If there are multiple event logs in the same region, the earliest event sequence is used.
- If there are event logs of different types in the same region, event sequence numbers for bad blocks caused by medium errors on RAID member drives take precedence.
- If a range of bad blocks runs across record boundaries, the sequence number corresponding to the last record is used.
The maximum number of dump files is 20.
An invocation example
dumpmdiskbadblocks 3
The resulting output if the MDisk has bad blocks:
Cluster name: my_cluster
Timestamp of dump: Fri Oct 31 11:27:33 2009 UTC
Mdisk id: 3
Mdisk name: mdisk3
Number of bad blocks: 6
Start LBA: 0x1234123412341234
Length: 2
Event log sequence number: 1
Start LBA: 0x5678567812341234
Length: 4
Event log sequence number: 2
The resulting output if the MDisk has no bad blocks:
Cluster name: my_cluster
Timestamp of dump: Fri Oct 31 11:27:33 2009 UTC
Mdisk id: 3
Mdisk name: mdisk3
Number of bad blocks: 0