I have an ESX 4 U2 server on a DL360 on which I recently increased the size of the raid array by adding two drives. Now, the ESX host sees the new space but I can't use it to extend the VMFS volume because it is in an extended partition. I think I need to add the new space to the extended partition first and then extend the VMFS volume to use it.
My question is - what is the best way to extend the extended partition in VMware to include the new space without losing any data that is already on it?
Anyone have any thoughts on this? I need to extend the 'Extended Partition' that the VMFS partition or volume is a part of so that I can then extend the VMFS volume. What is the best way to do this?
The answer to this question depends on a couple of things:
If you have extended the array that contains the ESX install then you will not be able to extend the VMFS partition to use this space.
However if this is a different array than the ESX install you will be able to do either of the following:
- Expand the VMFS Datastore to use all the space
- Add an extent to the current VMFS Datastore
My preference is the expansion of the VMFS Datastore rather than creating another extent. I have tested and you can expand an extent - say that five times fast!!! Obviously I would backup in both causes but should work without losing any data. Again in test (and have done in production) this works without losing data however we did have a good backup before.
Glen, I have extended the Array and I can see in vCenter the free space that is now available. However, this free space is 'outside' the extended partition that holds the VMFS partition. So I think I need to extend the Extended partition to incorporate the free space. Then in vCenter, I believe I will be able to extend the VMFS volume to include it. I found KB 1007026 which is a little confusing but I am going to give it a try using the Sourceforge tool systemrescuecd-x86-1.6.3.iso and see if it will work.
Yes I too found the same article, however also found this article: http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=100912...
This one allows you to grow the VMFS datastore. I tested this on an ESXi server (4.1) and it worked - your mileage may vary and ensure you have a good backup. My configuration was as follows:
- local disk 12GB in size with a 11GB VMFS datastore called esx01_localvmfs_01
- second disk with 40GB added as extent to esx01_localvmfs_01
I then increased the size of the second disk to 45GB and could grow the datastore. Increased the size of the first disk to 20GB (also containing ESXi) and could not grow using space from disk 1 plus the extent from disk 2. Also did this with VMs (not running) on the datastore.
Followed the process in the attached article and grew esx01_localvmfs_01 so it takes all the space on disk1.
Thanks. I'll save that for next time. I ended up just deleting everything and re-creating the host as it was faster in the end. I am still unclear how you can delete partitions, then recreate them and have the data on those partitions stay in tact.
It is because you don't write the partition table out between deleting and creating. If after deleting the partition, you press w then the partition would have been written and the partition would have been gone.
Trust this makes sense.