• This topic has 7 replies, 4 voices, and was last updated August 3, 2020 by Kalsang D.

Changed provisioning type while failover

  • Hi

    I wonder why when we failover an eager zero provisioned VM, it does not just replicate itself, but rather changes to a lazy zero.

    It was SQL VM and has 4 hard disks which are all set eager zeroed at first but changed to lazy zero.

    Hi Hyunwoo,

    This is Quincy from Zerto support team.

    Zerto components by default creating the recovery disk in lazy zero thick provisioning type.

    There is no option for changing this default behavior.

    If the recovered VM in the future failover live should have disks in eager zero thick type, please kindly refer to the steps we discussed in your previous service request.

    Thank you.

    Hi,

    Can you share the steps?

    “If the recovered VM in the future failover live should have disks in eager zero thick type, please kindly refer to the steps we discussed in your previous service request.”

     

    Thank you

    Hi PS M,

    The VPG has to be removed with keeping recovery disks and then recreated by using the kept disks as preseeding volumes, after the provisioning type of the disks is converted.

    Before recreating the VPG, the kept recovery disks should be converted to Thick Eager Zero provisioning.

    Please kindly refer to the following more detailed steps:

    1. Back up the VPG’s details;

    2. Remove the VPG with Keep Recovery Disks option selected;

    3. Change the kept recovery disks to Thick Eager Zero provisioning by running the following command on the relevant host:

    vmkfstools –eagerzero /vmfs/volumes/DatastoreName/VMName/VMName.vmdk

    Note: Please use the correct path of the recovery disks.

    More details can be found in the following VMware KB: https://kb.vmware.com/s/article/1035823

    4. Recreate the VPG by using the converted recovery disks as preseeding volumes.

    Thank you.

    Hi Quincy

    Can you tell me why I should recreate VPG?

    Can’t original VPG take settings of disk type if source VM’s disk type would changed?

    Hi Hyunwoo,

    Thank you for your reply.

    For changing the provisioning type of the recovery disk, you will have to remove the VM from the VPG, or remove the VPG itself when the VPG has only one VM or all VM in the VPGs are involved.

    This is for fully updating the VPG’s settings which are relevant to the disk provisioning type.

    Without readding the VM or recreating the VPG, there won’t be the preseeding process.

    If you directly change the recovery datastore settings in the VPG, without removing the VPG or readding the VM, the VPG won’t be able to update the provisioning information of the disk.

    Thank you.

    Hi Quincy

    Thanks for reply

    I found a log that a VPG which were not recreated had updated itself when I change a disk type of source VM
    (Lazy-Zeroed or Thin -> Eager-Zeroed)

    I think this is not a formal process you suggest but it was updated definitely without recreate VPG

    Thank you for the information. Please let us know If you’ve any other questions.

    Best regards,

The forum ‘Support Q & A’ is closed to new topics and replies.