cancel
Showing results for 
Search instead for 
Did you mean: 
Brent_Blood_102
Historic F5 Account

Problem this snippet solves:

This iCall script is a demo showing how a perpetual iCall handler can be used to watch for changes to the contents of the directory used to store core files (created when BIG-IP daemons unexpectedly exit), and upon seeing a new or modified entry, collecting a diagnostic snapshot of the system.

How to use this snippet:

Implementation Details

This iCall script requires v11.4 or higher.

Code :

### start code ###
sys icall script core_restart_watch {
    app-service none
    definition {
        # prime the array with existing cores so the initial run won't generate
        # qkviews
        foreach { file } [glob /var/core/*.gz] {
            file stat $file corestat
            set cores($file) $corestat(mtime)
        }

        # main loop
        while { 1 } {
            foreach { file } [glob /var/core/*.gz] {
                file stat $file corestat
                # act if an entry doesn't already exist (indicating new core),
                # or if one exists but has a different modification time
                # (indicating an overwritten corefile)
                if { ! [info exists cores($file)] || $cores($file) ne $corestat(mtime) } {
                    # insert/update this core for future use
                    set cores($file) $corestat(mtime)

                    # name the qkview after the hostname and date+time
                    set date [clock format [clock seconds] -format "%Y%m%d%H%M%S"]
                    set settings [tmsh::get_config sys global-settings]
                    set host [tmsh::get_field_value [lindex $settings 0] hostname]

                    # generate a UCS at this time
                    puts "core found ($file), collecting diagnostic data as: /var/tmp/${host}-${date}.ucs"
                    exec tmsh save sys config file /var/tmp/${host}-${date}.ucs
                    
                    # no need to run for more than one core
                    break
                }
            }
            # time in ms to wait between this running
            after 30000
        }
    }
    description none
    events none
}

### end code ###

And the following perpetual handler to run it:

### start code ###
handler perpetual core_restart_watch 
sys icall handler perpetual core_restart_watch {
    script core_restart_watch
}
### end code ###
Version history
Last update:
‎09-Mar-2015 15:49
Updated by:
Contributors