Unsolved
This post is more than 5 years old
2 Posts
0
26760
Seek help from PERC4 experts, about replacing controller...
Dear all,
Our old PE2850, which is running linux, is suddenly down, and when reboot, the onboard PERC4 controller said memory / battery has problem. The OS cannot be entered anymore.
Then I borrowed a PCI-X PERC4/D and hope can read back the volume and perform back up.
I have performed the following:
0. removed the hard disks from the machine
1. Plug in the PCI-X Raid card
2. Disabled the onboard PERC4 raid
3. Plug back the disks
4. Enter the Raid setting
5. In view configuration section, I select read from Disk, then save using this configuration, as the configuration seems to be the previous one in the onboard raid.
6. When reboot, the RAID volume seems to be found.
Then my questions are:
0. Are the onboard and the PCI-X one compatible?
1. Is the above procedure, especially step 5, is correct.
2. When I remove the disks from the machine, we forgot to record the disk order. We just plug back the disks randomly in step 3. Will this affect the volume, especially the data integrity, even the raid 5 volume could be recognized?
Thanks. Any advice is welcomed.
Patrick
DELL-Chris H
Moderator
Moderator
•
8.4K Posts
0
May 28th, 2013 13:00
Patrick,
The error you received is generally due to the ROMB dimm failing, not the controller. What you can do is power down the system and then remove the ROMB battery cable to the controller, then reseat the dimm, then restore the battery. If that doesn't work you can remove battery and dimm and then install dimm in the memory slot for the server (with other dimms removed) and see if the dimm recognizes, that will tell us if the dimm is failed. If so you may just need to replace the dimm and nothing else.
**NOTE: The ROMB battery MUST be removed prior to insertingthe ROMB dimm, as it makes the slot powered and can damage the dimm if installed.***
With the move over to the PCI-x controller, the process should work. If you pull the config from the drives the locations won't matter as it will know what drive goes where in the array.
Let me know how it goes.
lau0001
2 Posts
0
May 28th, 2013 23:00
Thanks a lot. While the raid volume seems to be recognized by the PCI-X card, and its status is shown to be optimal, when booting to the rescue mode of linux, the partition could not be mounted, saying something ext3 group descriptor corrupted. If my procedure is correct, then may be the file system is corrupted due to the raid card related (e.g. memory / battery) errors. Will try to rescue the disk by other methods.