Attachment 154206 Details for Bug 227420 – A shell script to monitor for and attempt to fix a sky2 driver wedge

A shell script to monitor for and attempt to fix a sky2 driver wedge

sky2mon (text/plain), 2.09 KB, created by Douglas Needham on 2007-05-05 15:34:21 UTC

(hide)

Description:

Filename:

MIME Type:

Creator: Douglas Needham

Created: 2007-05-05 15:34:21 UTC

Size: 2.09 KB

patch

obsolete

>#!/bin/ksh
>#
>#  Copyright (c) 2007, Douglas Needham  (ka8zrt.com)
>#  All rights reserved.
>#
>#  This file is freely redistributable under the terms of the NetBSD
>#  Foundation Copyright/License, which can be found at
>#
>#	http://www.netbsd.org/Goals/redistribution
>#  
>set -x
>echo $$ > /var/run/sky2mon.pid
>
>#
>#  Local variables
>#
>INTERVAL=60
>PROBES=(192.168.0.254 192.168.0.1 192.168.0.2)
>
>fix_sky2 ()
>{
>    #
>    #  Compute elapsed time.
>    #
>    now=`date +%s`
>    elapsed=$((now-last_time))
>
>    #
>    #  Log the restart
>    #
>    logger -p daemon.crit -t sky2mon -is "sky2 driver hang detected (${loop_count} iterations, ${elapsed} seconds): reloading driver and restarting network"
>
>    #
>    #  Stop the network
>    #
>    /etc/init.d/network stop
>
>    #
>    #  Show the module state
>    #
>    lsmod | grep sky2
>
>    #
>    #  Remove the sky2 driver
>    #
>    rmmod sky2
>    sleep 1
>
>    #
>    #  Show the module state
>    #
>    lsmod | grep sky2
>
>    #
>    #  Load the driver
>    #
>    modprobe sky2
>    lsmod | grep sky2
>
>    #
>    #  Restart the network
>    #
>    /etc/init.d/network restart
>
>    #
>    #  Reset last_time and loop_count
>    #
>    loop_count=0
>    last_time=`date +%s`
>}
>
>#
>#  Initialize loop counter and time
>#
>loop_count=0
>last_time=`date +%s`
>
>#
>#  Loop
>#
>while [ 1 ]
>do
>    #
>    #  Make sure that the local interface is up
>    #
>    ping -n -q -c 3 127.0.0.1
>    if [ $? -ne 0 ]
>    then
>	printf "Local network interface is down... sleeping"		>&2
>	sleep 300
>	continue
>    fi
>
>    #
>    # Start probing known hosts
>    #
>    down=0
>    idx=0
>    while [ ${idx} -lt ${#PROBES[*]} ]
>    do
>
>	#
>	#  Probe the host
>	#
>	ping -n -q -c 3 ${PROBES[$((idx++))]}
>
>	#
>	#  Did it respond?
>	#
>	if [ $? -eq 0 ]
>	then
>
>	    #
>	    #  It did... our network is working... no need to continue.
>	    #
>	    down=0
>	    break;
>        fi
>        down=$((down+1))
>    done
>
>    #
>    #  Were probed hosts unaccessable?
>    #
>    if [ ${down} -ne 0 ]
>    then
>	fix_sky2
>    fi
>
>    #
>    #  Increment our count and sleep for our interval
>    #
>    echo "Finished with iteration $((++loop_count))...sleeping"
>    sleep ${INTERVAL}
>done

Actions: View

Attachments on bug 227420: 154206 | 154207