[j-nsp] email from commit or op script?

Richard A Steenbergen ras at e-gerbil.net
Wed Feb 17 13:19:17 EST 2010


On Wed, Feb 17, 2010 at 10:42:18AM -0500, Phil Shafer wrote:
> To move in another direction: Has anyone added [transfer-on-commit]
> support to RANCID?

There isn't really a way to do that, both because rancid is a "pull the
config down via an expect script" model, and because it actually tracks
a lot more than just the config. For example, from the jrancid source:

        {'show chassis clocks'          => 'ShowChassisClocks'},
        {'show chassis environment'     => 'ShowChassisEnvironment'},
        {'show chassis firmware'        => 'ShowChassisFirmware'},
        {'show chassis fpc detail'      => 'ShowChassisFpcDetail'},
        {'show chassis hardware detail' => 'ShowChassisHardware'},
        {'show chassis routing-engine'  => 'ShowChassisRoutingEngine'},
        {'show chassis scb'             => 'ShowChassisSCB'},
        {'show chassis sfm detail'      => 'ShowChassisSCB'},
        {'show chassis ssb'             => 'ShowChassisSCB'},
        {'show chassis feb detail'      => 'ShowChassisSCB'},
        {'show chassis feb'             => 'ShowChassisSCB'},
        {'show chassis cfeb'            => 'ShowChassisSCB'},   
        {'show chassis alarms'          => 'ShowChassisAlarms'},
        {'show system autoinstallation status' => 'ShowSystemAutoinstall'},
        {'show system license'          => 'ShowSystemLicense'},
        {'show system license keys'     => 'ShowSystemLicenseKeys'},
        {'show system boot-messages'    => 'ShowSystemBootMessages'},
        {'show system core-dumps'       => 'ShowSystemCoreDumps'},
        {'show version detail'          => 'ShowVersion'},
        {'show configuration'           => 'ShowConfiguration'}

Of course the reason most people use rancid is for its simplicity (well, 
more or less :P) and its ability to support multiple platforms from 
multiple vendors. I'm sure you COULD whip up an op script replacement 
for it, I'm just not sure you'd find an extensive user base for it given 
those considerations.

On Wed, Feb 17, 2010 at 07:39:04AM -0600, Bill Marquette wrote:
> The device in question is an SRX and it's feasible (and probable given 
> our change windows) that multiple technicians could be making changes
> on it in rapid fire

Have you ever tried committing on an SRX? :) Don't worry, there is no 
danger of anything that could even be remotely described as "rapid fire" 
happening there. :P

-- 
Richard A Steenbergen <ras at e-gerbil.net>       http://www.e-gerbil.net/ras
GPG Key ID: 0xF8B12CBC (7535 7F59 8204 ED1F CC1C 53AF 4C41 5ECA F8B1 2CBC)


More information about the juniper-nsp mailing list