Chkdsk and hardware RAID

I have come across some articles stating that chkdsk should not be run a
server with hardware RAID.

Is this correct ?

Please post the articles. They are not correct imho.
"John gordon" <> wrote in message
news:...
>I have come across some articles stating that chkdsk should not be run a
>server with hardware RAID.
>
> Is this correct ?

Not that I know of. Where did you see those articles?

--
Claus
"John gordon" <> wrote in message
news:...
>I have come across some articles stating that chkdsk should not be run a
>server with hardware RAID.
>
> Is this correct ?

Along with the others I'd like to see those articles also.

I did find an article that talks about a 2002 report of Microsoft Issuing a
report on CHKDSK on a RAID Volume
http://www.lionbridge.com/NR/rdonlyr...2/mschkdsk.pdf

I doubt if they would pay for this evaluation if it wasn't recommended.

Russ

--

SBITS.Biz
Microsoft Gold Certified Partner
Microsoft Certified Small Business Specialist.
MCP, MCPS, MCNPS, (MCP-SBS)
North America Remote SBS2003 Support - http://www.SBITS.Biz


-

"John gordon" <> wrote in message
news:...
>I have come across some articles stating that chkdsk should not be run a
>server with hardware RAID.
>
> Is this correct ?

Can't recall if either of these were the ecact article that I originally saw
but

see http://www.quetek.com/RAID_mistakes.htm and
http://www.dataclinic.co.uk/raid-server-faults.htm

also one or two discussion threads elsewhere with differing opinions.
Personally chkdsk fixed a problem for me on a mirrored array recently with
no problem at all, although it did also once go wild on a Dell RAID 5 array
replacing thousands of security descriptors , although I don't think that
was anything to with hardware RAID.




"WorkingHard" <> wrote in message
news:...
> Please post the articles. They are not correct imho.
> "John gordon" <> wrote in message
> news:...
>>I have come across some articles stating that chkdsk should not be run a
>>server with hardware RAID.
>>
>> Is this correct ?
>
>

The first article says not to do this "while raid recovery is in process",
not in genreal.


"John gordon" <> wrote in message
news:%23aEL$...
> Can't recall if either of these were the ecact article that I originally
> saw but
>
> see http://www.quetek.com/RAID_mistakes.htm and
> http://www.dataclinic.co.uk/raid-server-faults.htm
>
> also one or two discussion threads elsewhere with differing opinions.
> Personally chkdsk fixed a problem for me on a mirrored array recently with
> no problem at all, although it did also once go wild on a Dell RAID 5
> array replacing thousands of security descriptors , although I don't think
> that was anything to with hardware RAID.
>
>
>
>
> "WorkingHard" <> wrote in message
> news:...
>> Please post the articles. They are not correct imho.
>> "John gordon" <> wrote in message
>> news:...
>>>I have come across some articles stating that chkdsk should not be run a
>>>server with hardware RAID.
>>>
>>> Is this correct ?
>>
>>
>

this is utter garbage trying to scare you into buying their product.

http://www.quetek.com/RAID_mistakes.htm
----quote----
Chkdsk being run on one of the disks
Chkdsk is a Windows utility that uses simple algorithms to recover data from
a corrupt disk. Chkdsk can be started by the user or automatically by
Windows if it detects disk corruption at boot time. (Windows will ask for
confirmation before starting chkdsk but will start it anyway if no response
is received after a number of seconds.) Chkdsk is not RAID-aware. Its
algorithms only work in the simplest cases. In general, it must be avoided.

If chkdsk attempts to repair one or more member disks, the RAID data
patterns will be destroyed, usually beyond recovery.
----end quote----

CHKDSK of NTFS is not a simple process. We are talking about NTFS because
SBS forces NTFS to be used for a minimum of the OS partition and the
partition containing the Exchange database. (personally, any system that
supports NTFS is converted to it as soon as I become aware that it's not
using it)

Windows CHKDSK _will NOT_ run against a member of a software RAID array
unless that member is in some way mistaken as a standalone drive, a most
unusual circumstance. It is _much more likely_ that an error on the drive
would result in the drive being orphaned from the set and showing up as a
'foreign drive'.

Should Windows CHKDSK detect an error on a hardware based RAID array it will
perform _exactly_ the same as if the array were a single HDD. This is the
whole purpose of hardware RAID, to hide the underlying structure from the
OS.

Should Windows CHKDSK run against the remaining members of the array it is
likely to cause intense activity, forcing the RAID mechanism (either soft or
hard) to address any RAID problem before allowing CHKDSK access to the
effected data. It's gonna be slow, it may result in softRAID not rebuilding
due to error.

YES. There are circumstances where such _may_ occur. The site is not lieing.


I agree with the principles of both sites, you have to understand the
underlying cause. The primary concern is not what Windows does with the
drive while it is in a degraded state, and it's better to not let Windows at
it, you need to 1st get the array functional. Should you not be able to get
array functional just about anything you do is going to lessen their chance
of reclaiming data from whatever remains of the array. Should you or the
system force a CHKDSK while the array is in a degraded state it is less
likely they will be able to recover your data from whatever remains.


"wedor" <> wrote in message
news:...
> The first article says not to do this "while raid recovery is in process",
> not in genreal.
>
>
> "John gordon" <> wrote in message
> news:%23aEL$...
>> Can't recall if either of these were the ecact article that I originally
>> saw but
>>
>> see http://www.quetek.com/RAID_mistakes.htm and
>> http://www.dataclinic.co.uk/raid-server-faults.htm
>>
>> also one or two discussion threads elsewhere with differing opinions.
>> Personally chkdsk fixed a problem for me on a mirrored array recently
>> with no problem at all, although it did also once go wild on a Dell RAID
>> 5 array replacing thousands of security descriptors , although I don't
>> think that was anything to with hardware RAID.
>>
>>
>>
>>
>> "WorkingHard" <> wrote in message
>> news:...
>>> Please post the articles. They are not correct imho.
>>> "John gordon" <> wrote in message
>>> news:...
>>>>I have come across some articles stating that chkdsk should not be run a
>>>>server with hardware RAID.
>>>>
>>>> Is this correct ?
>>>
>>>
>>
>
>

Well those "specialists" are in my black book ... of BS sellers based on
the misinformation they publish.

Cheers

"John gordon" <> wrote in message
news:%23aEL$...
> Can't recall if either of these were the ecact article that I originally
> saw but
>
> see http://www.quetek.com/RAID_mistakes.htm and
> http://www.dataclinic.co.uk/raid-server-faults.htm
>
> also one or two discussion threads elsewhere with differing opinions.
> Personally chkdsk fixed a problem for me on a mirrored array recently with
> no problem at all, although it did also once go wild on a Dell RAID 5
> array replacing thousands of security descriptors , although I don't think
> that was anything to with hardware RAID.
>
>
>
>
> "WorkingHard" <> wrote in message
> news:...
>> Please post the articles. They are not correct imho.
>> "John gordon" <> wrote in message
>> news:...
>>>I have come across some articles stating that chkdsk should not be run a
>>>server with hardware RAID.
>>>
>>> Is this correct ?
>>
>>
>

I came across this thread through a Google search and see it's quite old, but thought I'd revive it since I saw the info on the aforementioned website and agree it is wrong, but it is also right.

They are wrong chkdsk cannot be performed on a RAID ARRAY. However I agree it SHOULD NOT be performed on a SINGLE (physical, that is) RAID disk.

In my case, I discovered this when I changed my disks from RAID (2 disks in RAID 0) to IDE in the BIOS with the intention of running WD Data Lifeguard on them individually (can't be run on the array). In this case, I would see 2 disks instead of one and be able to run the disk check within Windows. On reboot, Windows Vista sees one disk with a lot of garbage (half of the data from the array, right?) so it automatically (I should have halted it, but missed the chkdsk 'countdown' to abort) starts peforming chkdsk and starts deleting 'file segments'. When I saw this I immediately panicked and performed a hard reboot. On reboot I get a RAID array failure in the Intel BIOS on one of the disks in the array. Windows shows nothing in "computer" where the disk used to be. Intel Matrix Storage Manager lists drive 1 of the 2 drive RAID array as "missing." At this point, real panic sets in since this is a 3TB disk with 10+ years of music, videos, pictures, documents, etc.

The aforementioned website, says "If chkdsk attempts to repair one or more member disks, the RAID data patterns will be destroyed, usually beyond recovery." This is clearly a plug for their expensive services, since it took me about 5 minutes to fix this problem that nearly gave me a heart attack.

First, I had to rebuild the RAID array in the Intel BIOS. It lists 1 disk as part of the array and the other as missing. I remove the existing disk from the array and then rebuild exactly as it was before from the 2 remaining free disks. Warnings about loss of data can be ignored, this procedure will not destroy any data.

Then booting into Windows you have a working array that, as a disk, is viewed as needing to have a partition created. Don't create a parition or do anything else to the disk though!

Download this free software, TestDisk, from: TestDisk - CGSecurity. Then follow the instructions here: TestDisk Step By Step - CGSecurity. On reboot, you should see your disk as it used to be. I have only gone through a cursory glance through files, etc but, for me, it all appears to be intact. I guess this is because chkdsk is not actually deleting or altering 'files' as it is their entries in the partition table?? Honestly, I don't know the answer but just very thankful this guy has built such a simple piece of powerful software that is FREE! Perhaps a donation is warranted though.

Hope this information is useful to some people.