backup of x series chassis with vsx installed

[btone]

hi cpus,

i'm searching for ideas on how to backup complete vsx/crossbeam installations to minimize restore time (no "appl vsx install" / vsx patching etc.) in case the cpm hdds are completely destroyed.

steps:
1.) install base xos via install server
2.) restore cbs running-config
3.) install backup app
4.) restore tftpboot dirs

or:

1.) restore box via 3rd party backup solution with kind of netboot like the install server (possible with cbs hardware?) and then restore an image to the cpm hdds (problem with software raid?)

so far i talked to our backup guys - red hat el 3 is not supported anymore, so they will not do it.
secure shore nms is able to backup files and dirs - but complete vap-goup dirs? don't have a license to test that :-(

any ideas/tips/hints appreciated :-)

regards

btone

[woody]

Are you running single or dual CPMs in the chassis' ?

[cpcpc]

You should be able to push back VSX configuration from CHeck Point management. Which version of VSX are you using?

[wa1di]

With initial config of VSX on Crossbeam and working SmartCenter/Provider-1 machine you should be able to restore it very quickly. After reenabling SIC topology should be repushed to APM modules and you should be able to reinstall policy.

[btone]

Originally Posted by woody

Are you running single or dual CPMs in the chassis' ?

single cpm only :-( cb already told us to use dual cpm to minimize the risk of a cpm failure..

[btone]

Originally Posted by wa1di

With initial config of VSX on Crossbeam and working SmartCenter/Provider-1 machine you should be able to restore it very quickly. After reenabling SIC topology should be repushed to APM modules and you should be able to reinstall policy.

if a cpm complete crashes and you need to restore a blank one you need to
- install xos
- install app vsx
- patch vsx (hfa01, patches + kernel paramters)
- vsx_util reconfigure

but this will take approx. 2-4h (rough estimate) which is to long for the SLAs.

[cpcpc]

CP shouldn't crash. Using two CPs will save you a lot of time if the HD/database is corrupted...