how to check data fragmentation in SAN

[CA975438]

SAN is very similar to disk drives. I suspect data can get fragmented over a period of time. How can you check data framentation for Server drives attached to SAN. How Could you defrag data.

please help

[Eugeny Brychkov]

From OS perspective SAN drives are exactly the same as local direct attached disks. You can use the same tools you usually use - for example, in windows 2000, right click to SAN drive in explorer, choose properties, and then tools tab. You will find 'disk defragmenter' tool there.
Does your SAN experience performance problems?
Eugeny

[CA975438]

Yes

Our disk queuess are going very high. We have several databases sitting on the SAN. I am monitoring both Physical disk and Logical disk counters on our Win2k environment. Server is clustered. Question is, is it advisable to check defrag with win2k utility on a clustered server for a disk sitting on the SAN.

[Eugeny Brychkov]

What's hardware? What's SAN disk array? Are these queues were always high? Do you run special heavy operations loading disk array? Did you load it this way before and did you see such high queues?
I would not blame fragmentation... disk arrays a little different devices than ordinary disks (one big difference: they have a lot of cache).
It is possible that SAN is configured in non-optimal way
Eugeny

[CA975438]

See my replies under....thanx always

What's hardware? DL 760 8X,8 GBram

What's SAN disk array? - compaq storage works hsv 110

Are these queues were always high? - no off late since we added more stuff on the san

Do you run special heavy operations loading disk array? - do not understand the question however i can tell you that the queues are on account of heavy read write operations, same was going on earleir too however the queues were in control

Did you load it this way before and did you see such high queues? - do not understand the question

I would not blame fragmentation... disk arrays a little different devices than ordinary disks (one big difference: they have a lot of cache).

It is possible that SAN is configured in non-optimal way -could you elaborate as to what could be wrong.

Eugeny

[Eugeny Brychkov]

Unfortunately I'm not skilled in EVA configuration, but one thing I can see clearly. As soon as you were adding more hardware to the SAN, server now can see, poll and access more devices. But number of FC cards did not change. This means that increased number or transactions are going through the same number of HBAs. So maybe you should increase number of FC HBAs installed or, again, somehow configure 'load balancing' over installed cards (load balancing means, for example, if you have tapes and disk arrays you setup arrays are accessible through one HBA and tape devices through another HBA, or simply split LUNs set to two subsets and setup primary links to these sets through different HBAs).
Hope this helps
Eugeny