12-18-2012 08:54 AM
Hi All
Just wondering if anyone has seen this behaviour from coverage ?
Here is the story:-
We run NBU7504 with a hpux master.
Our master server was down at the weekend for OS patching and the 1st run of a coverage report for the master reported 2 different filesystems twice.
Here is a sample from the coverage report:-
/usr/openv /dev/vg01/openvlv YBSLUX06-B_BKUP
/usr/openv/netbackup/logs /dev/vg01/nbulogslv YBSLUX06-B_BKUP
/usr/openv/netbackup/logs /dev/vg01/nbulogslv YBSLUX06-B_BKUP
/var /dev/vg00/lvol8 YBSLUX06-B_BKUP
/var/opt/perf /dev/vg01/perflv YBSLUX06-B_BKUP
/var/opt/perf /dev/vg01/perflv YBSLUX06-B_BKUP
Here is a sample of the bpmount command:-
bpmount -i
MNT - VxFS 19 12 /dev/vg01/scratchlv /var/scratch
MNT - VxFS 16 13 /dev/vg01/perflv /var/opt/perf
MNT - VxFS 16 13 /dev/vg01/perflv /var/opt/perf
MNT - VxFS 15 4 /dev/vg00/lvol7 /usr
MNT - VxFS 17 10 /dev/vg01/openvlv /usr/openv
MNT - VxFS 19 25 /dev/vg01/nbulogslv /usr/openv/netbackup/logs
MNT - VxFS 19 25 /dev/vg01/nbulogslv /usr/openv/netbackup/logs
MNT - VxFS 18 7 /dev/vg01/tivolilv /tivoli
Has anyone come accross this or have an explanation.?
Many thanx
12-18-2012 09:30 AM
Can you show a df -k output from the client too?
12-18-2012 09:38 AM
Here are the parts of interest I hope.
/tmp (/dev/vg00/lvol4 ) : 10163272 total allocated Kb
9847088 free allocated Kb
316184 used allocated Kb
4 % allocation used
/usr/openv/netbackup/logs (/dev/vg01/nbulogslv ) : 18358400 total allocated Kb
11438984 free allocated Kb
6919416 used allocated Kb
38 % allocation used
/usr/openv (/dev/vg01/openvlv ) : 15378664 total allocated Kb
2948072 free allocated Kb
12430592 used allocated Kb
81 % allocation used
/usr (/dev/vg00/lvol7 ) : 7386144 total allocated Kb
4568128 free allocated Kb
2818016 used allocated Kb
39 % allocation used
Also I have checked the policy and both duplicated entries on the report only exist in the policy.
12-18-2012 09:39 AM
sorry
Also I have checked the policy and both duplicated entries on the report only exist once in the policy.
12-18-2012 09:55 AM
Paul,
I don't think it's anything to do with the policy as running bpmount -i on the client shows the duplicate results for those 2 filesystems and bpmount -i does not check policies. I wonder if this is a Volume Manager issue or a bpmount issue.
Can you check if there is a /usr/openv/netbackup/include_list file on the client with those duplicate filesystems listed?
If not you could try using tusc on the bpmount to see what it's doing. You may need to raise a call with Symantec if it's causing issues for you.
12-18-2012 10:04 AM
... and it happened directly after OS patches were added ...
Were any of teh patches related to filesystems ?
Martin
12-18-2012 10:57 AM
That's a good point mph999, I missed that the client is also the master, I guess the clue was in the separate filesystem /usr/openv/netbackup/logs/ . A list of the patches and what they patch may be useful for trouble shooting this.
12-19-2012 05:03 AM
Hi and thank you for your input,
re - /usr/openv/netbackup/include_list there is no file on the system (now systems involved see below)
However I have had a chance to dig deeper and found that there we have other servers showing a similar behavior.
Different filesystems are involved also.
All effected servers are patched to 7504.
All servers still at 656 awaiting upgrade are not effected.
We also have 2 servers that have not had the OS patches that are still at 656 (currently not effected)
I have upgraded one of these and the symptom appeared - Must be 75 causing this.
Raising a call.
12-19-2012 05:23 AM
Must be NBU 7.5 with HP-UX combination. Not seeing the same on RHEL with 7.5.0.4.
01-07-2013 07:58 AM
Hi all,
Just posting a quick update,
We also have the symptoms on our Linux clients (Actually worse as "/" is reported 4 times) and it appears after both upgrades and new installs.
However I recently during a fresh install I made the following discovery, The symptoms only appeared when we placed our exclude_list file into the /usr/openv/netbackup directory.
When I remove the exclude list the symptoms disappear.
Looks like something new in the bpmount command perhaps interrogating the exclude_list ?
01-08-2013 12:31 AM
Looks like something new in the bpmount command perhaps interrogating the exclude_list ?
Well it certainly does look there: http://www.symantec.com/business/support/index?page=content&id=TECH141363
Do the entries in the exclude_list correspond directly to the duplications you're seeing in bpmount -i ?
01-08-2013 02:30 AM
Hi Andy,
Re 'Do the entries in the exclude_list correspond directly to the duplications you're seeing in bpmount -i ?`
No -
Here is our standard Linux exclude list:-
/tmp
/proc
/dev
/mnt
/var/tmp
/sys
And here is a snapshot of the coverage report from the same Linux box:-
Mount Point Device Backed Up By Policy Notes
----------- ------ ------------------- -----
/ /dev/mapper/vg00-lvol1 SERVER_BKUP
/ /dev/mapper/vg00-lvol1 SERVER_BKUP
/ /dev/mapper/vg00-lvol1 SERVER_BKUP
/ /dev/mapper/vg00-lvol1 SERVER_BKUP
/boot /dev/mapper/mpath0p1 SERVER_BKUP
/boot /dev/mapper/mpath0p1 SERVER_BKUP
/esm /dev/mapper/vg00-esmlv SERVER_BKUP
/home /dev/mapper/vg00-lvol6 SERVER_BKUP
/home /dev/mapper/vg00-lvol6 SERVER_BKUP
/install_master /dev/mapper/vg01-imlv SERVER_BKUP
/install_master /dev/mapper/vg01-imlv SERVER_BKUP
01-08-2013 02:42 AM
'Do the entries in the exclude_list correspond directly to the duplications you're seeing in bpmount -i ?`
That was a *really stupid* question of mine to say the least! Why would there be a correlation - one is an exclude the other essentially an include - DOH!
The only correlation I see is that there are 6 entries in the exclude & a 'corresponding' number of duplications in bpmount - there must be something in the coding that is replacing the entries from the exclude with the duplications in the bpmount .... hopefully it is only cosmetic .....