Skip to main content

NetApp_Insight_2020.png 

NetApp Knowledgebase

FAQ: FlexVol Volumes

Views:
434
Visibility:
Public
Votes:
0
Category:
data-ontap-8
Specialty:
core
Last Updated:

 

Applies to

  • Data ONTAP 7
  • Data ONTAP 8
  • ONTAP 9

Answer

  1. When you create a flexible volume, where is the volume created?

    The volume is created in the aggregate you specified. For example:
    vol create flex_vol_1 aggr_1 30m
    creates flex_vol_1 in aggregate aggr_1
     
  2. When you run aggr status does it tell you which flex volumes the aggregate contains?

    aggr status will not tell you this information, but aggr status <aggr_name> gives more detailed information.
     
  3. How often do we delete aggregate level snapshots?

    There is a snapshot schedule that contains hourly and nightly snapshots. Aside from the deletes caused by those, we may delete snapshots if the aggregate is running low on space. This is entirely load dependent.
     
  4. If aggregate-level snapshot auto-delete is ON, but no-snap=off is set, what snapshots are we deleting?

    The aggregate level snapshots are created for internal purposes. Whenever we delete those snapshots, we delete enough of them to free up space.
     
  5. How does aggregate-level snapshot delete mechanism know when to 'stop' deleting snapshots?

    It 'doesn't'. It frees things up until there is sufficient free space.
     
  6. When we autodelete the aggregate snapshots, are we left without any aggregate-level snapshots?

    No. We'll be left with least one aggregate-level snapshot when the delete is done. That is, we'll create a new one as part of this deletion.
     
  7. Why is snapshot auto-delete set to ON by default?

    This has an impact on space-accounting. In the absence of this option, the 2X space multiplier for overwritten files applies at both the aggregate and the volume levels, totaling to a 4X multiplier.
     
  8. Can we turn off space reservations on an aggregate?

    No. Space guarantees are a per-volume option and space reservations are within volumes.
     
  9. What are the uses for aggregate-level snapshots?

    Aggregate-level SnapRestore, aggr copy, syncmirror resync.
     
  10. Can you recover a flexible volume that you accidentally deleted?

    Maybe. If you have an aggregate snapshot which has that flexible volume in it, you can snap restore the aggregate. This will affect all other volumes.
    For another restore option, see 3012449: Is it possible to restore an individual flexible volume using an aggregate snapshot?
     
  11. Are aggregates, by default, created with RAID-DP?

    Yes.
     
  12. What does aggr create [-L] do?

    It is used for creating a snaplock aggregate.
     
  13. Does changing raidtype of an aggregate from RAID-DP to RAID4 free up the 2nd parity disk?

    Yes. This is the same as for a traditional volume.
     
  14. Are there any snap commands for the aggregate?

    Yes. The volume snap commands are applicable for an aggregate. They are accessed via the "-A" option  on the snap commands.
     
  15. For SnapMirror transfers, does the RAID type matter?

    No. Flexible volume snapmirror transfers are above the level of RAID and thus are completely independent of the RAID configuration and geometry.
     
  16. What is snap reclaimable?

    Displays the amount of space that can be reclaimed when the list of snapshots passed in (as input argument) is deleted.
     
  17. Why is guarantee-volume(disabled) set on a SnapMirror destination?

    Prior to Data ONTAP 8, quarantees were disabled by design. With guarantees disabled, it was possible for the destination aggregate to run out of space, thereby causing the SnapMirror transfers to fail. Data ONTAP 8 and later do not contain this behavior.
     
  18. Why can you QSM from a qtree on a traditional volume to a qtree on a flex volume, but can't SnapMirror from a trad volume to a flexible volume?

    QSM operates at the logical level (think files); hence, the type of volume doesn't matter. SnapMirror operates at block level (think disk blocks); the buftree internals of a trad/flex volume are different; hence, you can't move blocks between different volume types.
     
  19. What does guarantee=none mean on a flexible volume?

    This means that space for the flexible volume isn't guaranteed by the aggregate. Writes on the volume could fail if the aggregate containing the volume becomes full, even before the volume has used an space on the aggregate.
     
  20. With guarantee=none on a file/volume, can we get ENOSPC on a LUN?

    Yes.
     
  21. Can we enable space guarantees on a clone?

    No. This may be changed in a future release.
     
  22. Can you wafliron flexible volumes directly?

    Volume Granular Iron was introduced in ONTAP 9.5.  For previous versions Wafliron must be run at the aggregate level.  See 3011877: Overview of wafliron.
     
    WARNING: Do not run wafliron without first contacting NetApp Technical Support.
  23. How big can a flexible volume be sized?

    The maximum FlexVol size depends on the filer platform. The System Configuration Guide contains the appropriate information.
    Note: The maximum size provided is the raw size including the 10% reserve.
     
  24. When can you get duplicate FSID's?

    When you create a new volume, while another aggregate on the filer is offlined. It can also happen during an aggr copy or when splitting a syncmirrored aggregate. The filer automatically handles those cases.
    This also possibly occurs if SnapRestore is performed.
  25. How can you tell where the root volume is?

    From the vol status command, look in the Options column.
     
  26. What does coredumping do with respect to aggregates?

    The core file is still dumped to /etc/crash.
     
  27. What happens when root volume is corrupted and we need the core from /etc/crash?

    When you boot the filer, and there is no root volume, as part of the boot process, we will create AUTOROOT. If there is any core to be dumped, it gets dumped there.
     
  28. How do I tell what snapshots are 'locked' by clones?

    snap list includes this information.
     
  29. What does aggregate-level mirror-split do?

    Is similar to "mirror-split" of a traditional volume. On a split, from the mirror'ed disks, a mirror'ed aggregate comes into existence; also, any flexible volumes that were contained in the mirror'ed aggregate come into existence. These replicas are renamed to avoid name conflicts, and their FSIDs are also rewritten when the volumes are mounted.
     
  30. Can you vol copy/migrate flexible volumes?

    Yes.
     
  31. What does reallocate do on volumes?

    This is WAFL scan reallocate. In the presence of fragmented files, it will re-write some of the data to lay it out better on disk.
     
  32. Can you move a volume from one aggregate to another?

    You could by vol copy or snapmirror.
     
  33. Does vFiler migrate work on flexible volumes?

    Yes
     
  34. When a flexible volume is marked inconsistent, how can we troubleshoot?

    Contact NetApp Technical Support

             Refer to KB 2013507: AutoSupport message: WAFL INCONSISTENT for details.

  1. In maintanence mode, why do have vol_show_ops with an 's', but vol_remove_op without an 's'?

    vol_show_ops lists all the pending operations that have been queued up in maintanence mode; vol_remove_op removes the last operation that was enqueued.
     
  2. How do you know the actual space used by a LUN?

    For space-reserved LUNs: In a volume that contains no snapshots, a LUN uses its size, as reported by lun status and lun size.  In a volume that contains snapshots, the LUN will use at least that much space, and will eventually grow to twice that amount.  That additional space will be displayed in the reserved column of the df -r display (and it's included in the Used column).  However, the df command is volume granular, so if there are multiple LUNs in the volume, there is no general way of seeing how much of the Reserved space can be attributed to one LUN and how much to another.
     
    (There is a non-general way of seeing this:  from an NFS client, the du command, applied to the LUN in an NFS mount point, will report the number of blocks actually in the LUN; this is what it's costing in the Reserved column).
     
    We generally recommend that you assume that all LUNs will end up using the maximum size (2X+delta) and allow for this size from the beginning.  The amount of space actually used by the LUN is generally less interesting because we expect that the LUN will eventually grow to use the maximum, so you might as well provision for that from the beginning.  The lun maxsize command can be used to see how large a LUN can be created in the free space of a volume while obeying the 2X+delta rule.
     
  3. What is the maximum number of aggregate snapshots that can be created or maintained?

    10

Additional Information

Add your text here.

 

 

 

  • Was this article helpful?