Forum Discussion

Mark_Garringer's avatar
10 years ago

Automating BMR ISO creation

Background: I have a RHEL 6.6 Master server running 7.6.0.4. I have a dedicated Windows BMR server (Win 2k8 R2) that I use to create bootable ISOs. Generally speaking, I create the BMR job to a specific point in time (as opposed to using 'current'). I was able to find the commands I need to run on the Master to create the new Point in Time configs and prepare to restore so I could automate this process.

Challenge: I would like to be able to create bootable ISO images (as via using the bmrbsa.exe utility) but from the command line (or via a script) to automate the process. I cannot seem to find the right logs as I did with the config and prepare to restore. Does anyone have any tips on logging to try or bmrsrtadm.exe command line usage?

  • Not sure, but I would have expected just this on its own to fail:

    # bmrsrtadm -cli

    ...and I say that because no actions were specified, so it just quits.  The CLI based reporting of bad / incorrect / missing command line qualifiers / switches / arguments / parameters is pretty much non-existent, and any bad arguments and switches etc are only reported in the VxUL OID log 125... which means that getting the CLI constructs right involves a bit of trial and error.

    And the command above does not drop in to a tool as such.  After all, it is un-documented.  I was testing on Windows, and it does appear to work ok.

    .

    For a Unix/Linux test perhaps these steps might be a suitable test to prove whether similar functionality exists for Unix/Linux:

    1) Enable detailed bmrsrtadm logging:

    # vxlogcfg -a -p 51216 -o 125 -s DiagnosticLevel=6 -s DebugLevel=6

    2) Create an SRT using normal methods.

    3) List the SRTs:

    # bmrs -operation list -resource srt

    4) Use bmrsrtadm to delete the SRT, because the syntax for deleting an SRT is fairly straightforward:

    # bmrsrtadm -cli -task delete -srtname test01

    5) List the SRTs again... has it been deleted?

    # bmrs -operation list -resource srt

    6) Check the VxUL logs:

    # vxlogview -p 51216 -o 125 -t 00:05:00

    .

    I don't have a Unix/Linux system prep'd up to test on.  If someone else could test?  And specify their OS version and NetBackup version (would be good to know that too).      :)

  • Not sure, but I would have expected just this on its own to fail:

    # bmrsrtadm -cli

    ...and I say that because no actions were specified, so it just quits.  The CLI based reporting of bad / incorrect / missing command line qualifiers / switches / arguments / parameters is pretty much non-existent, and any bad arguments and switches etc are only reported in the VxUL OID log 125... which means that getting the CLI constructs right involves a bit of trial and error.

    And the command above does not drop in to a tool as such.  After all, it is un-documented.  I was testing on Windows, and it does appear to work ok.

    .

    For a Unix/Linux test perhaps these steps might be a suitable test to prove whether similar functionality exists for Unix/Linux:

    1) Enable detailed bmrsrtadm logging:

    # vxlogcfg -a -p 51216 -o 125 -s DiagnosticLevel=6 -s DebugLevel=6

    2) Create an SRT using normal methods.

    3) List the SRTs:

    # bmrs -operation list -resource srt

    4) Use bmrsrtadm to delete the SRT, because the syntax for deleting an SRT is fairly straightforward:

    # bmrsrtadm -cli -task delete -srtname test01

    5) List the SRTs again... has it been deleted?

    # bmrs -operation list -resource srt

    6) Check the VxUL logs:

    # vxlogview -p 51216 -o 125 -t 00:05:00

    .

    I don't have a Unix/Linux system prep'd up to test on.  If someone else could test?  And specify their OS version and NetBackup version (would be good to know that too).      :)

  • sdo - thanks, there are some steps in the right direction here but ultimately I have not been able to get this working in my 7.6.0.4 environment. A few of the commands are different than what you listed from your testing and logging, but the idea is sound. I turned up the logging on a bunch of OIDs and got some good information, but there still seems to be a missing part where it combines the restore configuration with the ISO that is not being logged at all.