We have also done this many, many times on Windows File Systems.
The only time I ran into a problem was when I extended a heavy IO Clustered (MSCS) file share from 300GB to 500GB using the same procedure AranH outlined.
The lun went offline and Windows would not recognize it until the cluster was brought down. This actually happened to us twice and Microsoft actually owned up to a known issue and had us put on the latest SP.
Even though I've done this many times without incidence on several other systems, I've learned my lesson only perform these changes before or after work ¿ for highly visible systems and systems with High IO.
you can expand the lun online... you also need to expand the filesystem on windows server. this can be done using diskpart utility on windows... you will not need downtime for this..
I have done this on multiple systems: MCS clusters running SQL, File servers using DiskXtender to attach to a Centera, and stand alone file servers. I have never had to stop application i/o or schedule any downtime for the expansion of the LUN or the extending of the partition within the OS.
After the LUN expansion is complete, in Microsoft Disk Management you will see the physical disk as being larger and the partition will not occupy the entire space of the disk.
Run diskpart and select the volume (partition) that you want to expand (run "list vol"). Once the volume is selected run "extend" and you are done. If you go back into Disk Management you will see that the partition now occupies the entire physical disk.
Good point Bowling, this should definitely be done after business hours, as should any changes to a production environment that could impact business.
Bowling, What was the SP level you were at when you had the issues? I have done the same procedure on MCS Clusters at SP1 and have not had the problem, although it was after the peak processing hours on the SQL database. I was wondering if I had just been lucky or if it was SP1 that fixed the issue, sine SP1 updated diskpart.
Bowling1
97 Posts
0
July 10th, 2007 08:00
The only time I ran into a problem was when I extended a heavy IO Clustered (MSCS) file share from 300GB to 500GB using the same procedure AranH outlined.
The lun went offline and Windows would not recognize it until the cluster was brought down. This actually happened to us twice and Microsoft actually owned up to a known issue and had us put on the latest SP.
Even though I've done this many times without incidence on several other systems, I've learned my lesson only perform these changes before or after work ¿ for highly visible systems and systems with High IO.
Kiran3
410 Posts
1
July 10th, 2007 02:00
you also need to expand the filesystem on windows server. this can be done using diskpart utility on windows...
you will not need downtime for this..
Crasto
11 Posts
0
July 10th, 2007 03:00
Do I need to stop I/O request temporarily (by taking database server down ) during LUN expansion or during diskpart ?
Also any problem for cluster environment ?
Thanks & Regards
Walter
dynamox
9 Legend
•
20.4K Posts
1
July 10th, 2007 05:00
AranH1
2.2K Posts
0
July 10th, 2007 08:00
After the LUN expansion is complete, in Microsoft Disk Management you will see the physical disk as being larger and the partition will not occupy the entire space of the disk.
Run diskpart and select the volume (partition) that you want to expand (run "list vol"). Once the volume is selected run "extend" and you are done. If you go back into Disk Management you will see that the partition now occupies the entire physical disk.
AranH1
2.2K Posts
0
July 10th, 2007 09:00
Bowling,
What was the SP level you were at when you had the issues? I have done the same procedure on MCS Clusters at SP1 and have not had the problem, although it was after the peak processing hours on the SQL database. I was wondering if I had just been lucky or if it was SP1 that fixed the issue, sine SP1 updated diskpart.
Crasto
11 Posts
0
July 17th, 2007 02:00
Thanks a lot for all your suggestions.
I have successfully expanded the LUN during off hours & with small down time during which we took SQL service offline & expanded the LUN .
Thanks & Regards
Walter.