This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Deleting savepoints not clearing space on RDA repo

Hello,

I have vRanger backing up to a Dell Powervault DR4100 repository via RDA (Dell Rapid Data Access). We recently ran out of space so I went to the Repository View and completely removed ALL backup savepoints from a few VM's. It was at least 2 TB worth of unique data..  Over the next few days I checked my DR4100 to see how much space it would clear up and the available space hasn't increased a single byte.   The DR4100's cleanup process runs normally and finds nothing to clean.

It almost seems like vRanger removes the savepoint from its database (and thus not viewable in the Repository) but does NOT actually delete the files on the repository!  The space does not get reclaimed until I delete the RDA container on the DR4100, then suddenly all that space is available.  Unfortunately, I cannot delete this particular container. 

Normally if this were an NFS or CIFS container I would be able to view the files and maybe delete them manually, but that does not seem to be the case with RDA. Just wondering if anyone has come across this issue before or knows of a way of browsing an RDA container to see the files, or have vRanger REALLY delete the files. 

 

THank you~

  • It sound like multi-faceted problem:

    1. Ranger or DR device do not warn about depleted free space.
    2. There is doubt if deleting savepoint through Ranger GUI actually deletes corresponding files from RDA container.
    3. RDA container turns read-only after reaching some threshold (93-95% usage). It requires DR support rep intervention to turn container back to read-write.
    4. Ranger RDA repo (basically RDA container) is not browseable.

    First of all I have to mention that I do not support DR device and do not have much experience. As for the problems above I can say following:

    1. I believe there should be some configurable alert on DR device that will send email about free space available or threshold reached.
    2. There is a popular reposync script (on the same Ranger forum) that may help a bit. For example, you deleted one or more savepoints from GUI. You can run script and resync repo content with DB. If SP pops up in GUI, you are right. If not, you can double check script log for entries related to particular SP(s).
    3. Not much to do. DR Support reps do their job, they possess necessary skills.
    4. reposync script (as of current state) can be used to browse 1st and 2nd level dirs in the repo. It won't display files though, only dir names.
      If I am not mistaken, you use 2nd option after repo is selected from the list and confirmed accessible.
      Script may not work with most recent DR firmware.

    For reference, repo looks like this:

    root

        |- GlobalManifest.metadata

        |- VM (with GUID) dir

              |- Savepoint set dir (either single Full SP or Full with children Incr/Diff)

                         |- Savepoint(s), each consists of 2 metadata files + var file(s)

  • Thank you so much for the great response and for the tool (v3.2)! I ran the tool with the sync DB option and the logs show no entries for the VM SP I deleted, nor do they show up in the GUI. So I guess that means the fault is on the DR4100 side. Thanks again!
  • Even though it may not be obvious but DR admin can setup a system alert to warn about space utilization reaching some critical point. I would strongly recommend to use this feature.
    You do it in the console. Run command:
    system --storage --set_usage_alert 90
    where 90 is a percentage

    Run command to verify:
    system --show
    Output will contain line "Storage Usage Alert : 90%"