[BALUG-Admin] done: some (mostly non-production) BALUG services to have short outage this PM

[Michael Paoli]

and done, bit under 6 minutes outage.
Slightly longer than anticipated: found and fixed one minor
config booboo before getting the balug VM back up again.

Also handled at same time (and higher priority - more critical
production) SF-LUG.

references/excerpts:

2013-05-25 03:00:56+00:00 up 208.96.15.254 80
2013-05-25 03:15:33+00:00 up --> down 208.96.15.254 80
2013-05-25 03:21:28+00:00 down --> up 208.96.15.254 80

$ expand -t 4 < balug/ck_balug-sf-lug-v2.balug.org.
#!/bin/sh
exec ck_IP_tcpPORT 208.96.15.254 80
$ expand -t 4 < ~/bin/ck_IP_tcpPORT
#!/bin/sh
[ $# -eq 2 ] || {
     1>&2 echo "usage: $0 IP_address TCP_port"
     exit 1
}
IP="$1"
PORT="$2"
sleep=1
timeout=5
status=
while :
do
     if nc -w "$timeout" -z "$IP" 80; then
         # up
         case x"$status" in
             xup)
                 # no change in status
                 :
             ;;
             xdown)
                 # down --> up
                 echo $(TZ=GMT0 date +'%Y-%m-%d %H:%M:%S+00:00')  
"$status --> up $IP 80"
             ;;
             x)
                 # ??? --> up
                 echo $(TZ=GMT0 date +'%Y-%m-%d %H:%M:%S+00:00') "up $IP 80"
             ;;
         esac
         status=up
     else
         # timeout/down/refused
         case x"$status" in
             xdown)
                 # no change in status
                 :
             ;;
             xup)
                 # up --> down
                 echo $(TZ=GMT0 date +'%Y-%m-%d %H:%M:%S+00:00')  
"$status --> down $IP 80"
             ;;
             x)
                 # ??? --> down
                 echo $(TZ=GMT0 date +'%Y-%m-%d %H:%M:%S+00:00') "down $IP 80"
             ;;
         esac
         status=down
     fi
     sleep "$sleep"
done
$

http://lists.balug.org/pipermail/balug-admin-balug.org/2013-May/000707.html
http://linuxmafia.com/pipermail/sf-lug/2013q2/010038.html