HI there,
We have a 3 server ESX3.5 Cluster backed up by a HP StorageWorks MVA F-SCSI SAN.
A LUN has been mounted as /dev/sdb, originally sitting at 500Gb. We have since extended this LUN to 1Tb, however can't work out how to get ESX to use this extra space...
If i run fdisk -l, it gives
Disk /dev/sdb: 1099.5 GB, 1099511627776 bytes
255 heads, 63 sectors/track, 133674 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Device Boot Start End Blocks Id System
/dev/sdb1 1 65270 524281211 fb Unknown
This shows that it sees the extra space, but cant use it. I found kittara's thread here, however when i follow the instructions through,ESX mounts it as a completey seperate datastore...
The other thing i've looked at is under Datastore properties, there is an "Add EXtent" button. I clicked that, and selected the additional space on the lun, however there's a warning there:
The Current Disk layout will be destroyed. All file systems and data will be permanently lost.
Is this true? Will it completely delete all VM data currently on that LUN?
Is there anyway i can extend the existing datastore???
Any info appreicated.
cheers
Gavin
The key to that KB article is in the quote -"f you look in Target Identifier, you see the vmhbaW:X:Y:Z where the extent is to be placed. If your original disk was vmhba1:0:0:1 , that extent is created on vmhba1:0:0:2 ." the syntax is for the vmhbaw.x.y.z is w is the ESX Host HBA, x is the Storage Processor, y is the LUN and z is the partion on the LUN - in your case the storage is identified as vmhba0:1:1 so when you add the extent you will have the original VMFS datastore on vmhba0:1:1:1 and the extent will be on vmhba0:1:1:2 -
If you find this or any other answer useful please consider awarding points by marking the answer correct or helpful
Welcoem to the Forums - You can not dynamically extend a LUN from the storage side - the wat to grow a lUN is by adding an extent which means creating another LUN and in the properties of your VMFS datastore in the extents section adding another extent -
If you find this or any other answer useful please consider awarding points by marking the answer correct or helpful
OK, i found the following KB Article... The Add Extent wizard includes a warning message regarding data loss after extending a LUN.
However on the above ss, the Target Identifier appears to be the same as the existing Extent, so does that mean it will over-write all existing data?
Confused
Cheers
Gavin
The key to that KB article is in the quote -"f you look in Target Identifier, you see the vmhbaW:X:Y:Z where the extent is to be placed. If your original disk was vmhba1:0:0:1 , that extent is created on vmhba1:0:0:2 ." the syntax is for the vmhbaw.x.y.z is w is the ESX Host HBA, x is the Storage Processor, y is the LUN and z is the partion on the LUN - in your case the storage is identified as vmhba0:1:1 so when you add the extent you will have the original VMFS datastore on vmhba0:1:1:1 and the extent will be on vmhba0:1:1:2 -
If you find this or any other answer useful please consider awarding points by marking the answer correct or helpful
So what you're saying is that it will be added to a new extent on 0:1:1:2... Makes sense, and i think the attached screenshot backs it up...
I clicked past the Warning screen on the add extents, double checked the size and stopped at the complete screen - and it's showing an additional extent on 0:1:1:2...
So as one final check before i click complete, it's definately not going to remove the existing data, just format the new extent???
Sorry for asking again, just dont want to click the wrong thing and wipe everything out...
Cheers
Gavin
I dont really like using extents, it can become really confusing when multiple LUNs are used to make up one datastore, and I think it probably effects performance. If I was in this situation I would create a brand new LUN of 1TB then storage-vmotion the VMs onto the new LUN.
JoJo,
Cheers for the idea... Might be something to look at in the future, but for now, i need to get it working this way...
Cheers
Gavin
Ok, in the end i plucked up the courage and just went for it...
And it WORKED!!!
SO i now have an enlarged shared datastore, which all 3 nodes can now see...
Happy days...
Cheers for all the help..
Gavin
Glad it worked -
If you find this or any other answer useful please consider awarding points by marking the answer correct or helpful