Backup & Recovery

 View Only
  • 1.  Big number of hfscheck in VDP Integrity Check

    Posted Apr 04, 2019 09:59 AM

    Hello,

    VDP core system is sometimes down and I usually reboot the virtual appliance of VDP (with OS shutdown) because the server vendor reccomended us so at the first case.

    # Because of the error VDP:001 The most recent checkpoint for the VDP appliance is outdated

    Today after the lunch, I reboot it and executed integrty check of VDP manually. But it has not completed yet and it is only 5% progress now... (it is working, though)

    I found a hfscheck takes a time by status.dpn. I cannot find such a big number of "stripes" in any cases on Internet.

    What "Hfscheck" is? Why the number has been incresed? Should I just wait until it is complete? Please help if you have any information.

    ESXi version 6.5.0 Update2 Build #8935087

    vSphere Data Protection 6.1

    The result of #: status.dpn is below (the header removed):

    All reported states=(ONLINE), runlevels(fullaccess), modes=(mhpu+0hpu+0hpu)

    System-Status: ok

    Access-Status: full

    Last checkpoint: cp.20190404044644 finished Thu Apr 4 12:47:03 2019 after 00m 18s (OK)

    No GC yet

    Hfscheck in progress: started Thu Apr 4 12:48:04 2019 >> checked 896 of 14791 stripes (hfscheck)

    Maintenance windows scheduler capacity profile is active.

         WARNING: Scheduler is STOPPED.

         Next backup window start time: Fri Apr 5 21:00:00 2019 SGT

         Next maintenance window start time: Fri Apr 5 13:00:00 2019 SGT

    Thank you,



  • 2.  RE: Big number of hfscheck in VDP Integrity Check
    Best Answer

    Posted Apr 08, 2019 09:07 AM

    HFS Check = Hierarchical File System check.

    Integrity check consists of 2 parts

    - Checkpoint

    - HFS check

    Checkpoint is taking a state snapshot of the VDP appliance cp.20190404044644

    Your data in VDP is striped and stored across the data0? partitions.

    By default a checkpoint taken is invalid and is validated once the HFS Check is run against the latest checkpoint.

    HFS Checks are of 2 types

    - Rolling

    - Full HFS

    Rolling is incremental HFS check where only the modified stripes are validated

    Full HFS checks validates all the data stripes.

    If you run cplist, you can see what type of HFS Check is there for a checkpoint

    rol - rolling

    hfs - full

    Based up on this you can tell how many stripes are validated. In your case it is 14791 stripes need to be validated. Run cplist and you will know if this is incremental or full HFS Check

    Yes, please let the HFS check complete. It should not take too much of time to complete thought.

    # Suhas



  • 3.  RE: Big number of hfscheck in VDP Integrity Check

    Posted Apr 09, 2019 05:47 AM

    I understood about HFS well... Thank you for your kind response.

    The result of cplist is "rol" so it is saying there is such a big change in our servers every day..

    BTW Finally I can see "completed" after 18 hours and some minutes.

    I think it's the most important advice. Thanks!

    Yes, please let the HFS check complete. It should not take too much of time to complete thought.