Replaced disk - RAID Sync failed

  • Thread starter Thread starter Timo Lamm
  • Start date Start date
T

Timo Lamm

Hi,
I'm running a P4P800 Deluxe with 4 Maxtor 200GB disks as RAID 0+1. One disk
failed and was replaced by the same type. But the synchronization of this
new disk fails every time I try to start it (doesn't matter if I use the
controller bios or the Via raid tool). Synchronization of the other disk on
the same channel works properly.
Already changed the Master/Slave settings on the disks of the channel and
checked the cable.
I'm using the latest via drivers for the 6410 and the 1019 BIOS on the
board.

Any ideas?

Thanks in advance
Timo
 
This is where raid vendors need to splash out on some diagnostic s/w.
I suggest a google on raid 1 failures, but at the end of the day you can:
- guess based on google results,
- backup the files and rebuild the raid from scratch starting with a new
raid volume,
- hassle the vendor.

I would do all 3 and work on #2 pending results of #3.

This makes borked raid 1 #4 in 2 weeks.
 
You should be aware that a raid built with a specific manu.controler is
specific to that controler.
ie another manu raid controler may not even see the raid, if it does data
corruption / loss may result. Nore is it likely anothers raid diagnostics
software will diagnose any other than their own controllers raid
 
I am more than aware of this - there needs to be industry standardisation on
the storage / format of raid meta data with the objective that the drives
and raid configs become h/w vendor independant.
EG pull out a raid array from one manufacturers controller and plug itinto
anothers and have it work 100% - so long as the two controller implement the
same raid type and options.

The issues that seem to be arising are where a drive in a mirror fails, but
the 'good' drive that is left is not in fact good. An IBM OEM Adaptec
contoller failure recently complained of this and the solution provided by
IBM was a file level backup restore. I have also read 3 other failures that
sound quite similar, although the controllers offer no evidence. The
manifestation is failure to recreate the RAID 1 after some % of processing.

A good volume in a raid should be able to be scanned and volume format
errors corrected - meta data that is used by the controller being optionally
corrected. A bit like a chkdsk /f.

Obviously, every manufacturers controller diagnostics will be controller
specific. More info is needed.
 
Mercury said:
The issues that seem to be arising are where a drive in a mirror fails, but
the 'good' drive that is left is not in fact good. An IBM OEM Adaptec
contoller failure recently complained of this and the solution provided by
IBM was a file level backup restore. I have also read 3 other failures that
sound quite similar, although the controllers offer no evidence. The
manifestation is failure to recreate the RAID 1 after some % of processing.

A good volume in a raid should be able to be scanned and volume format
errors corrected - meta data that is used by the controller being optionally
corrected. A bit like a chkdsk /f.

Obviously, every manufacturers controller diagnostics will be controller
specific. More info is needed.
Click to expand...

Well, I've been able to recover a failed disk in a software RAID 1
array under Linux, but I've yet to be able to recover a failed disk in
an onboard RAID 1 array under WinXP. You don't always get what you
pay for!!
 
Back
Top