vBulletin Search Engine Optimization
| |||||||
| Register | FAQ | Members List | Calendar | Search | Today's Posts | Mark Forums Read |
| ||||
| E3500 running Solaris 8 with dual A1000s attached. A1000s are mirrored using DiskSuite. One LUN ( shown below using Group 1 ) had 4 drives: $> ./drivutil -d syddb621-a1000b Drives in Group for syddb621-a1000b Group Drive List [Channel,Id] Group 1: [1,0]; [2,0]; [1,1]; [2,1]; Group 2: [1,2]; [2,2]; Group 3: [1,4]; [2,4]; Group 4: [2,3]; Group 5: [1,5]; drivutil succeeded! $> ./drivutil -I syddb621-a1000b Group Information for syddb621-a1000b Group No. of RAID No. of Total Remaining LUNs Level Drives Space(MB) Space(MB) 1 1 0 4 68938 0 2 1 0 2 69385 0 3 1 0 2 34469 0 4 1 0 1 17234 0 5 1 0 1 17234 0 drivutil succeeded! $> ./drivutil -l syddb621-a1000b Logical Unit Information for syddb621-a1000b LUN Group Device RAID Capacity Status Name Level (MB) 0 1 c4t5d0 0 68938 Optimal 1 2 c4t5d1 0 69385 Optimal 2 3 c4t5d2 0 34469 Optimal 3 4 c4t5d3 0 17234 Optimal 4 5 c4t5d4 0 17234 Optimal drivutil succeeded! LUN 0 above is used as a submirror of metadevice, as shown below, as metadevice d25 ( with d15 as a submirror on the other A1000 ): $> metastat -p ....... d5 -m d15 d25 1 d15 1 1 c1t5d0s1 d25 1 1 c4t5d0s1 Furthermore, soft partitions are layered __ON TOP OF__ the d5 metadevice, like so: $> metastat -p ...... d101 -p d5 -o 1 -b 25165824 d102 -p d5 -o 25165826 -b 16777216 d103 -p d5 -o 41943043 -b 16777216 Now the scenario is: One of the drives in the LUN was replaced. Out of the drives [1,0]; [2,0]; [1,1]; [2,1] ... [2,1] was replaced with a new disk. The d25 submirror was detached. Then d25 was re-attached to the d5 metadevice. After that, metastat shows that d25 was being resynced. Resyncing has been completed ... Now wy question now is: Do I still have to run metarecover ( -d or -m ), or does the fact that the soft partitions are on top of a mirrored meta device mean that the soft partitions exten headers are synced as well ? The link below: http://sysunconfig.net/unixtips/soft-partitions.html .... under the section "Recovering Soft Partitions" says: 2. reads through the current replica and creates the soft partitions on the given slice. This is good to run after a disk fails and gets replaced with a new one. The option to use on the metarecover command is -m. That section of the article above does not cover the case of replacing a disk that has a soft partition on top of a mirrored meta-device on the replace disk. Now since I actually replaced a disk, I was wondering if I still need to do the above. If I verify / list the soft partitions on top of the d5 metadevice, the output is shown below ( essentially saying there is no problem ). However, does it actually reads extent headers on BOTH sides of the mirror ? $> metarecover -v -n d5 -p Verifying on-disk structures on d5. The following extent headers were found on d5. Name Seq# Type Offset Length d101 0 ALLOC 0 25165825 d102 0 ALLOC 25165825 16777217 d103 0 ALLOC 41943042 16777217 NONE 0 END 135024639 1 NONE 0 FREE 58720259 76304380 Found 3 soft partition(s) on d5. Checking sequence numbers. d5: Validating soft partition metadb entries. Verifying device d101. d101: Status "Okay" is valid. Extent Number Virtual Offset Physical Offset Length 0 0 1 25165824 Verifying device d102. d102: Status "Okay" is valid. Extent Number Virtual Offset Physical Offset Length 0 0 25165826 16777216 Verifying device d103. d103: Status "Okay" is valid. Extent Number Virtual Offset Physical Offset Length 0 0 41943043 16777216 d5: Soft Partition metadb configuration is valid d5: Verifying metadb allocations match extent headers. Metadb extent: Name Seq# Type Offset Length d101 0 ALLOC 0 25165825 Extent header extent: Name Seq# Type Offset Length d101 0 ALLOC 0 25165825 Metadb extent: Name Seq# Type Offset Length d102 0 ALLOC 25165825 16777217 Extent header extent: Name Seq# Type Offset Length d102 0 ALLOC 25165825 16777217 Metadb extent: Name Seq# Type Offset Length d103 0 ALLOC 41943042 16777217 Extent header extent: Name Seq# Type Offset Length d103 0 ALLOC 41943042 16777217 d5: Soft Partition metadb matches extent header configuration Any insight appreciated. Thanks |
| Thread Tools | |
| Display Modes | |
|
|