IBM ServeRAID, Справочник пользователя

Страница: 177 / 226

b. In certain circumstances, the ServeRAID defunct drive event log does not
list the drives in the exact order that the drives failed. For example, when an
array spans multiple ServeRAID channels, the ServeRAID controller issues
parallel I/O requests to devices on each channel. In the event of a
catastrophic failure, physical drives might be marked defunct at the same
time.
7. Remove the first physical drive marked defunct from the backplane or
disconnect the SCSI cable from the drive.
8. Exit ServeRAID Manager; then, shut down the server.
Checking the hardware connections
While the server is turned off, complete the following tasks:
1. Reseat the ServeRAID controllers.
2. Reseat the SCSI cables and the disks against the backplanes.
3. Reseat the power cables to the backplane and SCSI backplane repeater
options, if they are present.
As you are reseating the components, visually inspect each piece for bent pins,
nicks, crimps, pinches, or other signs of damage. Take extra time to ensure that
each component snaps or clicks into place properly.
Setting defunct drives to the online state
Complete the following steps to set the defunct drives to the online state:
1. Start the server.
2. Insert the IBM ServeRAID Support CD into the CD-ROM drive.
3. Use ServeRAID Manager to delete all hot-spare and standby-hot-spare drives.
This will prevent a rebuild operation from starting.
4. Set each defunct physical drive in the failed array to the online state. The failed
logical drives should change to a critical state. If there are problems bringing
physical drives online, or if a physical drive initially goes online and then fails to
a defunct state soon after, see “Troubleshooting” on page 162. The logical
drives must be in a critical state before you proceed.
Accessing the critical logical drives
Complete the following steps to attempt to access the critical logical drives:
1. If you are still in ServeRAID Manager, exit and restart the server.
2. Attempt to start the operating system. If you are prompted to perform any
file-system integrity tests, skip the tests.
3. If the operating system starts, run the IPSSEND getbst command on each
logical drive. This will determine if the bad stripe table has incremented on a
logical drive. If there are one or more bad stripes, data is lost or damaged.
4. Run a read-only, file-system integrity check on each logical drive.
5. Using application-specific tools, run more intensive integrity tests on each logical
drive.
6. If all tests pass, you can assume that the data is good. Go to “Finishing the
procedure” on page 162. Otherwise, weigh the results of the file-integrity tests
and determine how best to proceed. Your options include the following:
v Use a recent backup to restore damaged files.
v Use a recent backup to restore damaged files; schedule a planned outage to
rebuild the data at a later date.
Chapter 18. Solving ServeRAID problems
161