a way to verify duplication images are cleaned up

[D_Davis]

my client just asked me to verify that images on disk are deleting.

so first I checked the disk usage, then made sure all policies had an exp date. but now I want/need a command I can run to produce the information my client would need.

[mph999]

Any reason you think they are not ?

The images on the actual storage unit (ie. the actual backed up data) contain file names that contain the ctime, the same ctime that is part of the backupid.

There should be no image info in the catalog for images that have expired, well, at least only for a short time between the time they expire, and when a cleanup job runs that actually removes them.

I cannot think of any NBU command that can make the check that you require.

So, I would do this :

bpimagelist command to list the backupids ....

Use a script to compare these to the filenames on the disk

Eg.

[root@nbmaster00 netbackup]# bpimagelist -idonly -hoursago 48

Time: Tue 14 Jan 2014 03:10:15 PM PS   ID: nbmaster00_1389741015   FULL (0)

(bpimagelist shows 24hrs by default)

ls -al on disk ...

-rw-------  1 root root 32768 Jan 14 15:13 nbmaster00_1389741204_C1_F1.1389741204.img

-rw-------  1 root root  4096 Jan 14 15:13 nbmaster00_1389741204_C1_F1.1389741204.info

-rw-------  1 root root  4096 Jan 14 15:13 nbmaster00_1389741204_C1_HDR.1389741204.img

-rw-------  1 root root  4096 Jan 14 15:13 nbmaster00_1389741204_C1_HDR.1389741204.info

Then I'd use a script to compare the two ...

If on the other hand you just wish to check if there are expired images just in the imeage catalog, run nbcc, and log a call to get in analized.

M

[Marianne]

Please tell us more about this environment.

Which NBU version?

Are disk backups duplicated to tape?
If so, how? SLP or Disk Staging?

Can you see regular Image Cleanup jobs in Activity Monitor?

If we know more about the environment, we can point you to the appropriate documentation.

In the meantime, ensure that these log folders exist on all media servers:
admin bptm bpdm bpbrm

bpdm will log local disk cleanup.

[mph999]

Just to clarify ....

If there are only fragments on disk, and these do not exist in the catalog at all, then there is no command, you have to compare one with the other manually (ie. does this fragment I found on disk exist in the catalog).

If there are images in the catalog that have expired and also exist /not exist on tape, then nbcc is the way to go.

Providing you are not using AIR, the ctime that is part of the filename on disk/ part of the backupid is unique, so is an easy way to search. 

nbdelete -allvolumes -force will clean up expired fragments on disk, but only if information telling NBU theyy have expired exists in the catalog, or more acurately, the NBDB deleted fragents table.

[D_Davis]

Please tell us more about this environment:

Which NBU version? : its ver 7.1 run on windows 08 servers

Are disk backups duplicated to tape? Yes!

If so, how? SLP or Disk Staging? I am a novice to disk and am looking for a way to get this information. Storage lifecycle policies are setup so I do not think disk staging is used.

Can you see regular Image Cleanup jobs in Activity Monitor? Yes about 3 times per day , some complete with 1 others with 0.

as for is there any reason to think the images arent expiring? NO

but do they want me to explain and check and show? Yes. so I will comply.

but i wanted something more than just "hey look at the activity monitor" to validate this for them, as I am still trying to sell them long term on the validity of netbackup long term.

[Marianne]

Activity Monitor is a starting point.

Status 1 for Image Cleanup means that part of the Image Cleanup job has failed.

You need to look at Details Tab of this job to see which images failed to cleanup and why.
You then need to look at bpdm log of the relevant media server to troubleshoot.

If you are not 100% of duplication method, check if policies are going to a Storage Unit or a SLP. 
Check Schedules as well.
Check SLP to validate retentions for backup and duplications.
(nbsu ooutput on the master server will give you ALL config info in text files.)
Use methods documented in SLP Best Practice pdf to monitor SLP progress and ensure that backlogs do not occur. 

[D_Davis]

1/14/2014 9:00:55 PM - Critical bpdm(pid=1332) sts_open_server failed: error 2060014 operation aborted       
1/14/2014 9:00:55 PM - Critical bpdm(pid=1332) failure to open sts for storage server MY-nbu-mediaSrv: plug-in reports error 2060014 operation aborted

and every policy backups to a lifecycle

[Marianne]

As per my previous post - check bpdm log on MY-nbu-mediaSrv to troubleshoot error.

You clearly have images NOT cleaning up.