When planning for an upgrade from vSphere4 to vSphere5 what will be the effect of upgrading a 1TB VMFS3 datast

When planning for an upgrade from vSphere4 to vSphere5 what will be the effect of upgrading a 1TB VMFS3 datastore with a 2MB block size to VMFS5?

A.
Thick virtual disks can be expanded to a maximum of 2TB.

B.
Thick virtual disks can be expanded to a maximum of 512GB.

C.
Thin virtual disks can be expanded to a maximum of 512GB.

D.
Thin virtual disks can be expanded to a maximum of 2TB.

13 Comments on “When planning for an upgrade from vSphere4 to vSphere5 what will be the effect of upgrading a 1TB VMFS3 datast

  1. Mohsin Alvi says:

    Answer should be “C”

    as

    When upgrading a VMFS datastore from VMFS-3 to VMFS-5 you can extend a datastore past 2TB – 512B. The caveat to upgrading a VMFS-3 to VMFS-5 datastore is that it will inherit the block size properties of the original VMFS-3 datastore.
    If you originally had a 1TB VMFS-3 datastore with a block size of 2MB and you upgrade the datastore to VMFS-5 and extend it to 10TB the maximum size VMDK file will still be 512GB (which is the limit of a VMFS-3 2MB block size).
    If you want to create files larger then 512GB then you need to migrate the virtual machines off the existing datastore, delete the VMFS volume and recreate it as a new VMFS-5 datastore.

    http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1003565

      1. sam says:

        I did not pay attention to the word “virtual disks” which is the VMDK file, so really is no change in file size from the upgrade, but you can have more 512GB (max) vmdk created in the datastore. is that right?

    1. Mike says:

      Answer D is correct. With VMFS-5 the block size no longer determines the maximum size of a VMDK file. This is why the default block size of 1MB can be used to create 2TB disks.

      The previous VMFS-3 datastore could only support a maximum of 512GB VMDKs. When you upgrade, you end up with a VMFS-5 datastore that has a block size of 2MB. Since the new filesystem can support 2TB VMDKs regardless of block size, you will now be able to create a 2TB VMDK.

    1. Mohsin Alvi says:

      Its “c” @sam

      when you create thick disk, it occupied all the space up to defined during creation.
      but Thin disk grows as data fill (write operation) upto defined limit.

      1. sam says:

        thx for the reply, this question has been more cleared (at least to me), bcz it given the specific size (1TB VMFS3 datastore with a 2MB block),it is not a simple Max question.
        the term of expand is not refer to vdisk expansive, it is really the grows inside of the vmdk.
        the question is very tricky, it combines sugar and muster in it.
        happy VS.

        1. Prakash BS says:

          D is correct.

          A.INCORRECT.
          Thick disk cannot be expanded to 2TB cos datastore is only 1 TB
          B. INCORRECT
          Thick disk can be expanded more than 512GB.cos datastore has 1TB

          C.INCORRECT. Thin disk size can be bigger than datastore size.

          D.
          Thin virtual disks can be expanded to a maximum of 2TB.

  2. DCW says:

    Thanks Prakash, I’ve read this question 100 times and couldn’t figure out why Thin not Thick.

    No matter what version of VMFS you have, you cant have a Thick Provisioned disk larger than the datastore its on (so simple)

  3. Bharath says:

    When a VMFS-3 Volume with 2 MB Block size is upgraded to VMFS-5 which has 1 MB Block Size(By Default and only option) it still continues to use 2 MB block size till the datastore is re-formatted with VMFS5.

    FYI – When VMFS5 datastore is grown more than 2TB it uses GPT(no change in block size unless datastore is re-created) while VMFS-3 used MBR.

    Again, why only thin? Why can’t thick disks be grown till 512 Gb? I don’t understand!!!

    This question is very contradictory. Answer cannot be A or D as per my understanding.

  4. mkader says:

    When upgrading a VMFS datastore from VMFS-3 to VMFS-5, you can extend a datastore past “2TB minus 512B”. The caveat to upgrading a VMFS-3 datastore to VMFS-5 is that it will inherit the block size properties of the original VMFS-3 datastore.

    If you upgrade to VMFS-5 from VMFS-3 then regardless of the block size, VMFS-5 uses double-indirect addressing to cater for large files (up to a size of “2TB minus 512B”) on upgraded VMFS-3 volumes. For example, if the VMDK goes beyond 512 GB it will switch to using double-indirect addressing, which will allow for VMDKs up to “2TB minus 512B”.

    Three things to remember to answer this questoin:
    1. Thick provisioned VMDK can not be larger than the datastore it is on.
    2. Thin provisioned VMDK can be larger than the datastore it is on.
    3. Maximum size of VMDK (thick or thin) can be 2TB (VMFS-5 file system).

    A. NOT correct, just because the datastore is 1TB and hence thick disk can not be expanded more than 1TB.
    B. NOT Correct, just because with a datastore of 1TB, think provisioned vmdk can be upto 1TB.
    C. NOT Correct, max size of a thin provisioned VMDK can be 2TB, even if the datastore is 1TB in size.
    D. Correct, Even though the size of the datastore is 1TB, since thin provisioned VMDK can be larger than the datastore it resides, it can be expanded to a maximum size of 2TB which is the limit of VMDK in VMFS-5 file system.


Leave a Reply

Your email address will not be published. Required fields are marked *