Re: Raid 1 and 5 Chkdsk

Read this http://www.dataclinic.co.uk/raid-server-faults.htm

You should never run chkdsk in /F or /F/R mode on a RAID array. Chkdsk is
not RAID aware, and will simply try to analyse and fix the NTFS tree as if
dealing with a normal disk. In doing so, it can write over information in a
degraded RAID array, that might have been recoverable at a lower level. You
can run it in read mode, but I would use the RAID controllers software. You
should be able access during machine boot up process. This should have
several options on checking for disk failure and recovery.

Chkdsk might be able to fix file system errors but it cannot do a surface
scan since it does not have access to any of the physical surfaces. I have
lost an array doing chkdsk on a raid array. That is why proper power
backup, data backup and recovery procedures are such a critical part of any
server setup.


Bryan


"Brian Cryer" <not.here@localhost> wrote in message
news:...
> <> wrote in message
> news:c25a8ff9-7d7c-4e2c-bbcd-...
>> Hi all.
>>
>> I've googled a lot, but I've found out various and different opinions:
>>
>> Is it correct to use chkdsk /F and chkdsk /F /R on a RAID 1 system and
>> on a RAID 5 (with 3 HDs) system ?
>> All my systems are hardware RAIDs and the servers are domain
>> controller with w2k srv and w2k3 srv.
>
> I would have thought so, yes - because a RAID controller should present
> the RAID as a single disk to windows and windows shouldn't even need to be
> aware that its a RAID.
>
> chkdsk /f checks for logical corruptions, which could still occur whatever
> type of RAID you have if you ever shutdown in an ungraceful manner.
>
> chkdsk /r checks for bad sectors. This is probably pointless because even
> if one of the physical disks had a bad sector the raid controller (whether
> raid 1 or raid 5) should return the correct contents.
>
> So I would have thought chkdsk /f but not /r.
>
> Presumably if the RAID is failing (controller error or multiple disk
> errors) then chkdsk might report an error but chkdsk will not be able to
> fix the problem because the problem is with the underlying RAID. In this
> scenario chkdsk might even make things worse! So if you ever need to run
> chkdsk also check the state of the RAID controller.
> --
> Brian Cryer
> www.cryer.co.uk/brian
>
>
>

I have an MSI K9N SLI mb with a NVIDIA® nForce 570 SLI MCP Chipset. I have 3
disks installed, configured as Raid 5. (with 2 partitions)

The F10 Bios showed the raid as Healthy, however, I have just installed
Vista Ultimate 64 and the F10 Bios is now showing the raid as damaged. Two
disks are OK and one is in error - even though there is no hw problem.

At the start of the Windows install, I loaded the latest version of the
floppy raid and SATA drivers (even though the "raw" windows installation
could still see the partitions"). I have had a similar problem before when I
scheduled a chkdsk /f /r in Windows Visa 64, after the reboot, the raid was
shown as damaged. In this case I was able to rebuild the raid with
MediaShield.

To me it looks like during installation or during a chkdsk, Windows is not
using the proper Raid driver and hence not maintiaining the Raid structure
properly.

The only way I was able to "recover" the situation was to use the disk
management utility under windows to delete the "damaged" raid disk, the
re-add it and rebuild the raid. The system is back "working OK".

1) How are you supposed to install windows on a Raid 5 cluster and why,
after loading the driver at install time, did windows not maintain the raid 5
structure?

2) What can you do to keep the NTFS structure correct if every time you run
CHKDSK it corrupts the raid 5 array?