#!/bin/sh
#
# lustre   This shell script takes care of starting and stopping Lustre
#
# chkconfig: - 99 1
# description: Lustre Lite network File System.
#              This starts both Lustre client and server functions.
# processname: lconf
# config: /etc/lustre/config.xml
# pidfile: /var/run/lustre.pid
### BEGIN INIT INFO
# Provides: lustre
# Required-Start:
# Required-Stop: 
# Should-Start: scance
# Should-Stop: scance
# Default-Start: 
# Default-Stop: 0 1 2 6
# Short-Description: Lustre Lite network File System.
# Description: This starts both Lustre client and server functions.
### END INIT INFO


SERVICE=lustre
LOCK=/var/lock/subsys/$SERVICE

: ${LUSTRE_CFG:=/etc/lustre/lustre.cfg}
[ -f ${LUSTRE_CFG} ] && . ${LUSTRE_CFG}

: ${LUSTRE_CONFIG_XML:=/etc/lustre/config.xml}
: ${LCONF:=/usr/sbin/lconf}
: ${LCONF_START_ARGS:="${LUSTRE_CONFIG_XML}"}
: ${LCONF_STOP_ARGS:="--force --cleanup ${LUSTRE_CONFIG_XML}"}
: ${LCTL:=/usr/sbin/lctl}

# Source function library.
if [ -f /etc/init.d/functions ] ; then
   . /etc/init.d/functions
fi

# Source networking configuration.
if [ -f /etc/sysconfig/network ] ; then
   . /etc/sysconfig/network
fi

check_start_stop() {
	# Check that networking is up.
	[ "${NETWORKING}" = "no" ] && exit 0

	[ -x ${LCONF} -a -x ${LCTL} ] || exit 0

	[ -f ${LUSTRE_CONFIG_XML} ] || ( echo "unconfigured" && exit 0 )

	# Create /var/lustre directory 
	# This is used by snmp agent for checking lustre services
	#    status online/offline/online pending/offline pending.

	[ -d ${STATUS_DIR:=/var/lustre} ] || mkdir -p $STATUS_DIR
}
STATUS=${STATUS_DIR}/sysStatus

start() {
	check_start_stop
	echo -n "Starting $SERVICE: "
	if [ $UID -ne 0 ]; then
		echo "Lustre should be started as root"
		RETVAL=1
		return
	fi
	${LCONF} ${LCONF_START_ARGS}
	RETVAL=$?
	echo $SERVICE
	if [ $RETVAL -eq 0 ]; then
		touch $LOCK
		echo "online" >$STATUS
	else
		echo "online pending" >$STATUS
	fi
}

stop() {
	check_start_stop
	echo -n "Shutting down $SERVICE: "
	if [ $UID -ne 0 ]; then
		echo "Lustre should be stopped as root"
		RETVAL=1
		return
	fi
	${LCONF} ${LCONF_STOP_ARGS}
	RETVAL=$?
	echo $SERVICE
	rm -f $LOCK 
	if [ $RETVAL -eq 0 ]; then
		echo "offline" >$STATUS
	else
		echo "offline pending" >$STATUS
	fi
}

restart() {
	stop
	start
}

status() {
	STATE="stopped"
	egrep -q "libcfs|lvfs|portals" /proc/modules && STATE="loaded"

	# check for any routes - on a portals router this is the only thing
	[ "`cat /proc/sys/portals/routes 2> /dev/null`" ] && STATE="running"
	
	# check for any configured devices (may indicate partial startup)
	[ "`cat /proc/fs/lustre/devices 2> /dev/null`" ] && STATE="partial"

	# check for either a server or a client filesystem
	MDS="`ls /proc/fs/lustre/mds/*/recovery_status 2> /dev/null`"
	OST="`ls /proc/fs/lustre/mds/*/recovery_status 2> /dev/null`"
	LLITE="`ls /proc/fs/lustre/llite/fs* 2> /dev/null`"
	[ "$MDS" -o "$OST" -o "$LLITE" ] && STATE="running"

	# check for server disconnections 
	DISCON="`grep -v FULL /proc/fs/lustre/*c/*/*server_uuid 2> /dev/null`"
	[ "$DISCON" ] && STATE="disconnected"

	# check for servers in recovery
	[ "$MDS$OST" ] && grep -q RECOV $MDS $OST && STATE="recovery"

	[ "`dmesg | grep LBUG`" ] && STATE="LBUG"

	echo $STATE
}

# See how we were called.
case "$1" in
  start)
	start
	;;
  stop)
	stop
	;;
  restart)
	restart
	;;
  status)
	status $SERVICE
	;;
  *)
	echo "Usage: $SERVICE {start|stop|restart|status}"
	exit 1
esac

exit $RETVAL