cluster-glue 1.0.12-5 / usr / sbin / hb_report

#!/bin/sh

 # Copyright (C) 2007 Dejan Muhamedagic <dmuhamedagic@suse.de>
 # 
 # This program is free software; you can redistribute it and/or
 # modify it under the terms of the GNU General Public
 # License as published by the Free Software Foundation; either
 # version 2.1 of the License, or (at your option) any later version.
 # 
 # This software is distributed in the hope that it will be useful,
 # but WITHOUT ANY WARRANTY; without even the implied warranty of
 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 # General Public License for more details.
 # 
 # You should have received a copy of the GNU General Public
 # License along with this library; if not, write to the Free Software
 # Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
 #

. /usr/lib/ocf/lib/heartbeat/ocf-shellfuncs

HA_NOARCHBIN=/usr/share/cluster-glue

. $HA_NOARCHBIN/utillib.sh

unset LANG
export LC_ALL=POSIX

PROG=`basename $0`

# the default syslog facility is not (yet) exported by heartbeat
# to shell scripts
#
DEFAULT_HA_LOGFACILITY="daemon"
export DEFAULT_HA_LOGFACILITY
LOGD_CF=`findlogdcf /etc $HA_DIR`
export LOGD_CF

SSH_PASSWORD_NODES=""
: ${SSH_OPTS="-o StrictHostKeyChecking=no -o EscapeChar=none"}
LOG_PATTERNS="CRIT: ERROR:"
# PEINPUTS_PATT="peng.*PEngine Input stored"

# Important events
#
# Patterns format:
#	title	extended_regexp
# NB: don't use spaces in titles or regular expressions!
EVENT_PATTERNS="
membership	crmd.*ccm_event.*(NEW|LOST)|pcmk_peer_update.*(lost|memb):
quorum		crmd.*crm_update_quorum:.Updating.quorum.status|crmd.*ais.disp.*quorum.(lost|ac?quir)
pause		Process.pause.detected
resources	lrmd.*rsc:(start|stop)
stonith		crmd.*te_fence_node.*Exec|stonith-ng.*log_oper.*reboot|stonithd.*(requests|(Succeeded|Failed).to.STONITH|result=)
start_stop	Configuration.validated..Starting.heartbeat|Corosync.Cluster.Engine|Executive.Service.RELEASE|crm_shutdown:.Requesting.shutdown|pcmk_shutdown:.Shutdown.complete
"

init_tmpfiles

#
# the instance where user runs hb_report is the master
# the others are slaves
#
if [ x"$1" = x__slave ]; then
	SLAVE=1
fi

usage() {
	cat<<EOF
usage: hb_report -f {time|"cts:"testnum} [-t time]
       [-u user] [-X ssh-options] [-l file] [-n nodes] [-E files]
       [-p patt] [-L patt] [-e prog] [-MSDZAQVsvhd] [dest]

	-f time: time to start from or a CTS test number
	-t time: time to finish at (dflt: now)
	-d     : don't compress, but leave result in a directory
	-n nodes: node names for this cluster; this option is additive
	         (use either -n "a b" or -n a -n b)
	         if you run $PROG on the loghost or use autojoin,
	         it is highly recommended to set this option
	-u user: ssh user to access other nodes (dflt: empty, root, hacluster)
	-X ssh-options: extra ssh(1) options
	-l file: log file
	-E file: extra logs to collect; this option is additive
	         (dflt: /var/log/messages)
	-s     : sanitize the PE and CIB files
	-p patt: regular expression to match variables containing sensitive data;
	         this option is additive (dflt: "passw.*")
	-L patt: regular expression to match in log files for analysis;
	         this option is additive (dflt: $LOG_PATTERNS)
	-e prog: your favourite editor
	-Q     : don't run resource intensive operations (speed up)
	-M     : don't collect extra logs (/var/log/messages)
	-D     : don't invoke editor to write description
	-Z     : if destination directories exist, remove them instead of exiting
	         (this is default for CTS)
	-A     : this is an OpenAIS cluster
	-S     : single node operation; don't try to start report
	         collectors on other nodes
	-v     : increase verbosity
	-V     : print version
	dest   : report name (may include path where to store the report)
EOF

[ "$1" != short ] &&
	cat<<EOF

	. the multifile output is stored in a tarball {dest}.tar.bz2
	. the time specification is as in either Date::Parse or
	  Date::Manip, whatever you have installed; Date::Parse is
	  preferred
	. we try to figure where is the logfile; if we can't, please
	  clue us in ('-l')
	. we collect only one logfile and /var/log/messages; if you
	  have more than one logfile, then use '-E' option to supply
	  as many as you want ('-M' empties the list)

	Examples

	  hb_report -f 2pm report_1
	  hb_report -f "2007/9/5 12:30" -t "2007/9/5 14:00" report_2
	  hb_report -f 1:00 -t 3:00 -l /var/log/cluster/ha-debug report_3
	  hb_report -f "09sep07 2:00" -u hbadmin report_4
	  hb_report -f 18:00 -p "usern.*" -p "admin.*" report_5
	  hb_report -f cts:133 ctstest_133

	. WARNING . WARNING . WARNING . WARNING . WARNING . WARNING .

	  We won't sanitize the CIB and the peinputs files, because
	  that would make them useless when trying to reproduce the
	  PE behaviour. You may still choose to obliterate sensitive
	  information if you use the -s and -p options, but in that
	  case the support may be lacking as well. The logs and the
	  crm_mon, ccm_tool, and crm_verify output are *not* sanitized.

	  Additional system logs (/var/log/messages) are collected in
	  order to have a more complete report. If you don't want that
	  specify -M.

	  IT IS YOUR RESPONSIBILITY TO PROTECT THE DATA FROM EXPOSURE!
EOF
	exit
}
version() {
	echo "cluster-glue: 1.0.12 (0a7add1d9996b6d869d441da6c82fb7b8abcef4f)"
	exit
}
#
# these are "global" variables
#
setvarsanddefaults() {
	local now=`perl -e 'print time()'`
	# used by all
	DEST=""
	FROM_TIME=""
	CTS=""
	TO_TIME=0
	HA_LOG=""
	UNIQUE_MSG="Mark:HB_REPORT:$now"
	SANITIZE="passw.*"
	DO_SANITIZE=""
	FORCE_REMOVE_DEST=""
	COMPRESS="1"
	# logs to collect in addition
	# NB: they all have to be in syslog format
	#
	EXTRA_LOGS="/var/log/messages"
	# used only by the master
	NO_SSH=""
	SSH_USER=""
	TRY_SSH="root hacluster"
	SLAVEPIDS=""
	NO_DESCRIPTION="1"
	SKIP_LVL=0
	VERBOSITY=0
}
#
# caller may skip collecting information if the skip level
# exceeds the given value
#
skip_lvl() {
	[ $SKIP_LVL -ge $1 ]
}
chkname() {
	[ "$1" ] || usage short
	echo $1 | grep -qs '[^a-zA-Z0-9@_+=:.-]' &&
		fatal "$1 contains illegal characters"
}
set_dest() {
	# default DEST has already been set earlier (if the
	# argument is missing)
	if [ $# -eq 1 ]; then
		DEST=`basename $1`
		DESTDIR=`dirname $1`
	fi
	chkname $DEST
	if [ -z "$COMPRESS" -a -e "$DESTDIR/$DEST" ]; then
		if [ "$FORCE_REMOVE_DEST" -o "$CTS" ]; then
			rm -rf $DESTDIR/$DEST
		else
			fatal "destination directory $DESTDIR/$DEST exists, please cleanup or use -Z" 
		fi
	fi
}
chktime() {
	[ "$1" ] || fatal "bad time specification: $2 (try 'YYYY-M-D H:M:S') "
}
no_dir() {
	fatal "could not create the working directory $WORKDIR"
}
time2str() {
	perl -e "use POSIX; print strftime('%x %X',localtime($1));"
}
# try to figure out where pacemaker ... etc
get_pe_state_dir() {
	PE_STATE_DIR=`strings $CRM_DAEMON_DIR/pengine |
		awk 'NF==1&&/var\/lib\/.*pengine$/'`
	test -d "$PE_STATE_DIR"
}
get_cib_dir() {
	CIB_DIR=`strings $CRM_DAEMON_DIR/crmd |
		awk 'NF==1&&/var\/lib\/.*(cib|crm)$/'`
	test -d "$CIB_DIR"
}
get_pe_state_dir2() {
	# PE_STATE_DIR
	local localstatedir lastf
	localstatedir=`dirname $HA_VARLIB`
	lastf=$(2>/dev/null ls -rt `2>/dev/null find /var/lib -name pengine -type d |
		sed 's,$,/*.last,'` | tail -1)
	if [ -f "$lastf" ]; then
		PE_STATE_DIR=`dirname $lastf`
	else
		for p in pacemaker/pengine pengine heartbeat/pengine; do
			if [ -d $localstatedir/$p ]; then
				debug "setting PE_STATE_DIR to $localstatedir/$p"
				PE_STATE_DIR=$localstatedir/$p
				break
			fi
		done
	fi
}
get_cib_dir2() {
	# CIB
	# HA_VARLIB is normally set to {localstatedir}/heartbeat
	local localstatedir
	localstatedir=`dirname $HA_VARLIB`
	for p in pacemaker/cib heartbeat/crm; do
		if [ -f $localstatedir/$p/cib.xml ]; then
			debug "setting CIB_DIR to $localstatedir/$p"
			CIB_DIR=$localstatedir/$p
			break
		fi
	done
}
get_crm_daemon_dir() {
	# CRM_DAEMON_DIR
	local libdir p
	libdir=`dirname $HA_BIN`
	for p in pacemaker heartbeat; do
		if [ -x $libdir/$p/crmd ]; then
			debug "setting CRM_DAEMON_DIR to $libdir/$p"
			CRM_DAEMON_DIR=$libdir/$p
			return 0
		fi
	done
	return 1
}
get_crm_daemon_dir2() {
	# CRM_DAEMON_DIR again (brute force)
	local p d d2
	for p in /usr /usr/local /opt; do
		for d in libexec lib64 lib; do
			for d2 in pacemaker heartbeat; do
				if [ -x $p/$d/$d2/crmd ]; then
					debug "setting CRM_DAEMON_DIR to $p/$d/$d2"
					CRM_DAEMON_DIR=$p/$d/$d2
					break
				fi
			done
		done
	done
}
compatibility_pcmk() {
	get_crm_daemon_dir || get_crm_daemon_dir2
	if [ ! -d "$CRM_DAEMON_DIR" ]; then
		fatal "cannot find pacemaker daemon directory!"
	fi
	get_pe_state_dir || get_pe_state_dir2
	get_cib_dir || get_cib_dir2
	debug "setting PCMK_LIB to `dirname $CIB_DIR`"
	PCMK_LIB=`dirname $CIB_DIR`
	# PTEST
	PTEST=`echo_ptest_tool`
	export PE_STATE_DIR CIB_DIR CRM_DAEMON_DIR PCMK_LIB PTEST
}

#
# find log files
#
logmark() {
	logger -p $*
	debug "run: logger -p $*"
}
#
# first try syslog files, if none found then use the
# logfile/debugfile settings
#
findlog() {
	local logf=""
	if [ "$HA_LOGFACILITY" ]; then
		logf=`findmsg $UNIQUE_MSG | awk '{print $1}'`
	fi
	if [ -f "$logf" ]; then
		echo $logf
	else
		echo ${HA_DEBUGFILE:-$HA_LOGFILE}
		[ "${HA_DEBUGFILE:-$HA_LOGFILE}" ] &&
			debug "will try with ${HA_DEBUGFILE:-$HA_LOGFILE}"
	fi
}

#
# find log slices
#

find_decompressor() {
	if echo $1 | grep -qs 'xz$'; then
		echo "xz -dc"
	elif echo $1 | grep -qs 'bz2$'; then
		echo "bzip2 -dc"
	elif echo $1 | grep -qs 'gz$'; then
		echo "gzip -dc"
	else
		echo "cat"
	fi
}
#
# check if the log contains a piece of our segment
#
is_our_log() {
	local logf=$1
	local from_time=$2
	local to_time=$3

	local cat=`find_decompressor $logf`
	local first_time="`$cat $logf | head -10 | find_first_ts`"
	local last_time="`$cat $logf | tail -10 | tac | find_first_ts`"
	if [ x = "x$first_time" -o x = "x$last_time" ]; then
		return 0 # skip (empty log?)
	fi
	if [ $from_time -gt $last_time ]; then
		# we shouldn't get here anyway if the logs are in order
		return 2 # we're past good logs; exit
	fi
	if [ $from_time -ge $first_time ]; then
		return 3 # this is the last good log
	fi
	# have to go further back
	if [ $to_time -eq 0 -o $to_time -ge $first_time ]; then
		return 1 # include this log
	else
		return 0 # don't include this log
	fi
}
#
# go through archived logs (timewise backwards) and see if there
# are lines belonging to us
# (we rely on untouched log files, i.e. that modify time
# hasn't been changed)
#
arch_logs() {
	local next_log
	local logf=$1
	local from_time=$2
	local to_time=$3

	# look for files such as: ha-log-20090308 or
	# ha-log-20090308.gz (.bz2) or ha-log.0, etc
	ls -t $logf $logf*[0-9z] 2>/dev/null |
	while read next_log; do
		is_our_log $next_log $from_time $to_time
		case $? in
		0) ;;  # noop, continue
		1) echo $next_log  # include log and continue
			debug "found log $next_log"
			;;
		2) break;; # don't go through older logs!
		3) echo $next_log  # include log and continue
			debug "found log $next_log"
			break
			;; # don't go through older logs!
		esac
	done
}
#
# print part of the log
#
print_log() {
	local cat=`find_decompressor $1`
	$cat $1
}
print_logseg() {
	if test -x $HA_NOARCHBIN/print_logseg; then
		$HA_NOARCHBIN/print_logseg "$1" "$2" "$3"
		return
	fi

	local logf=$1
	local from_time=$2
	local to_time=$3
	local tmp sourcef

	# uncompress to a temp file (if necessary)
	local cat=`find_decompressor $logf`
	if [ "$cat" != "cat" ]; then
		tmp=`mktemp` ||
			fatal "disk full"
		add_tmpfiles $tmp
		$cat $logf > $tmp ||
			fatal "disk full"
		sourcef=$tmp
	else
		sourcef=$logf
		tmp=""
	fi

	if [ "$from_time" = 0 ]; then
		FROM_LINE=1
	else
		FROM_LINE=`findln_by_time $sourcef $from_time`
	fi
	if [ -z "$FROM_LINE" ]; then
		warning "couldn't find line for time $from_time; corrupt log file?"
		return
	fi

	TO_LINE=""
	if [ "$to_time" != 0 ]; then
		TO_LINE=`findln_by_time $sourcef $to_time`
		if [ -z "$TO_LINE" ]; then
			warning "couldn't find line for time $to_time; corrupt log file?"
			return
		fi
	fi
	dumplog $sourcef $FROM_LINE $TO_LINE
	debug "including segment [$FROM_LINE-$TO_LINE] from $logf"
}
#
# print some log info (important for crm history)
#
loginfo() {
	local logf=$1
	local fake=$2
	local nextpos=`python -c "f=open('$logf');f.seek(0,2);print f.tell()+1"`
	if [ "$fake" ]; then
		echo "synthetic:$logf $nextpos"
	else
		echo "$logf $nextpos"
	fi
}
#
# find log/set of logs which are interesting for us
#
dumplogset() {
	local logf=$1
	local from_time=$2
	local to_time=$3

	local logf_set=`arch_logs $logf $from_time $to_time`
	if [ x = "x$logf_set" ]; then
		return
	fi

	local num_logs=`echo "$logf_set" | wc -l`
	local oldest=`echo $logf_set | awk '{print $NF}'`
	local newest=`echo $logf_set | awk '{print $1}'`
	local mid_logfiles=`echo $logf_set | awk '{for(i=NF-1; i>1; i--) print $i}'`

	# the first logfile: from $from_time to $to_time (or end)
	# logfiles in the middle: all
	# the last logfile: from beginning to $to_time (or end)
	case $num_logs in
	1) print_logseg $newest $from_time $to_time;;
	*)
		print_logseg $oldest $from_time 0
		for f in $mid_logfiles; do
			print_log $f
			debug "including complete $f logfile"
		done
		print_logseg $newest 0 $to_time
	;;
	esac
}

#
# cts_findlogseg finds lines for the CTS test run (FROM_LINE and
# TO_LINE) and then extracts the timestamps to get FROM_TIME and
# TO_TIME
#
cts_findlogseg() {
	local testnum=$1
	local logf=$2
	if [ "x$logf" = "x" ]; then
		logf=`findmsg "Running test.*\[ *$testnum\]" | awk '{print $1}'`
	fi
	getstampproc=`find_getstampproc < $logf`
	export getstampproc # used by linetime

	FROM_LINE=`grep -n "Running test.*\[ *$testnum\]" $logf | tail -1 | sed 's/:.*//'`
	if [ -z "$FROM_LINE" ]; then
		warning "couldn't find line for CTS test $testnum; corrupt log file?"
		exit 1
	else
		FROM_TIME=`linetime $logf $FROM_LINE`
	fi
	TO_LINE=`grep -n "Running test.*\[ *$(($testnum+1))\]" $logf | tail -1 | sed 's/:.*//'`
	[ "$TO_LINE" -a $FROM_LINE -lt $TO_LINE ] ||
		TO_LINE=`wc -l < $logf`
	TO_TIME=`linetime $logf $TO_LINE`
	debug "including segment [$FROM_LINE-$TO_LINE] from $logf"
	dumplog $logf $FROM_LINE $TO_LINE
}

#
# this is how we pass environment to other hosts
#
dumpenv() {
	cat<<EOF
DEST=$DEST
FROM_TIME=$FROM_TIME
TO_TIME=$TO_TIME
USER_NODES="$USER_NODES"
NODES="$NODES"
MASTER_NODE="$MASTER_NODE"
HA_LOG=$HA_LOG
MASTER_IS_HOSTLOG=$MASTER_IS_HOSTLOG
UNIQUE_MSG=$UNIQUE_MSG
SANITIZE="$SANITIZE"
DO_SANITIZE="$DO_SANITIZE"
SKIP_LVL="$SKIP_LVL"
EXTRA_LOGS="$EXTRA_LOGS"
USER_CLUSTER_TYPE="$USER_CLUSTER_TYPE"
CONF="$CONF"
B_CONF="$B_CONF"
PACKAGES="$PACKAGES"
CORES_DIRS="$CORES_DIRS"
VERBOSITY="$VERBOSITY"
EOF
}
is_collector() {
	test "$SLAVE"
}
is_node() {
	test "$THIS_IS_NODE"
}
is_master() {
	! is_collector && test "$WE" = "$MASTER_NODE"
}
start_slave_collector() {
	local node=$1

	dumpenv |
	if [ "$node" = "$WE" ]; then
		debug "running: $LOCAL_SUDO hb_report __slave"
		$LOCAL_SUDO hb_report __slave
	else
		debug "running: ssh $SSH_OPTS $node \"$SUDO hb_report __slave"
		ssh $SSH_OPTS $node \
			"$SUDO hb_report __slave"
	fi | (cd $WORKDIR && tar xf -)
}

#
# does ssh work?
# and how
# test the provided ssh user
# or try to find a ssh user which works without password
# if ssh works without password, we can run the collector in the
# background and save some time
#
testsshconn() {
	ssh $SSH_OPTS -T -o Batchmode=yes $1 true 2>/dev/null
}
findsshuser() {
	local n u rc
	local ssh_s ssh_user="__undef" try_user_list
	if [ -z "$SSH_USER" ]; then
		try_user_list="__default $TRY_SSH"
	else
		try_user_list="$SSH_USER"
	fi
	for n in $NODES; do
		rc=1
		[ "$n" = "$WE" ] && continue
		for u in $try_user_list; do
			if [ "$u" != '__default' ]; then
				ssh_s=$u@$n
			else
				ssh_s=$n
			fi
			if testsshconn $ssh_s; then
				debug "ssh $ssh_s OK"
				ssh_user="$u"
				try_user_list="$u" # we support just one user
				rc=0
				break
			else
				debug "ssh $ssh_s failed"
			fi
		done
		[ $rc = 1 ] &&
			SSH_PASSWORD_NODES="$SSH_PASSWORD_NODES $n"
	done
	if [ -n "$SSH_PASSWORD_NODES" ]; then
		warning "passwordless ssh to node(s) $SSH_PASSWORD_NODES does not work"
	fi

	if [ "$ssh_user" = "__undef" ]; then
		return 1
	fi
	if [ "$ssh_user" != "__default" ]; then
		SSH_USER=$ssh_user
	fi
	return 0
}
node_needs_pwd() {
	local n
	for n in $SSH_PASSWORD_NODES; do
		[ "$n" = "$1" ] && return 0
	done
	return 1
}
say_ssh_user() {
	if [ -n "$SSH_USER" ]; then
		echo $SSH_USER
	else
		echo your user
	fi
}

#
# the usual stuff
#
getbacktraces() {
	local f bf flist
	flist=$(
		for f in `find_files "$CORES_DIRS" $1 $2`; do
			bf=`basename $f`
			test `expr match $bf core` -gt 0 &&
				echo $f
		done)
	[ "$flist" ] && {
		getbt $flist > $3
		debug "found backtraces: $flist"
	}
}
pe2dot() {
	local pef=`basename $1`
	local dotf=`basename $pef .bz2`.dot
	test -z "$PTEST" && return
	(
	cd `dirname $1`
	$PTEST -D $dotf -x $pef >/dev/null 2>&1
	)
}
getpeinputs() {
	local pe_dir flist
	local f
	pe_dir=$PE_STATE_DIR
	debug "looking for PE files in $pe_dir"
	flist=$(
		find_files $pe_dir $1 $2 | grep -v '[.]last$'
	)
	[ "$flist" ] && {
		mkdir $3/`basename $pe_dir`
		(
		cd $3/`basename $pe_dir`
		for f in $flist; do
			ln -s $f
		done
		)
		debug "found `echo $flist | wc -w` pengine input files in $pe_dir"
	}
	if [ `echo $flist | wc -w` -le 20 ]; then
		for f in $flist; do
			skip_lvl 1 || pe2dot $3/`basename $pe_dir`/`basename $f`
		done
	else
		debug "too many PE inputs to create dot files"
	fi
}
getratraces() {
	local trace_dir flist
	local f
	trace_dir=$HA_VARLIB/trace_ra
	test -d "$trace_dir" || return 0
	debug "looking for RA trace files in $trace_dir"
	flist=$(find_files $trace_dir $1 $2 | sed "s,`dirname $trace_dir`/,,g")
	[ "$flist" ] && {
		tar -cf - -C `dirname $trace_dir` $flist | tar -xf - -C $3
		debug "found `echo $flist | wc -w` RA trace files in $trace_dir"
	}
}
touch_DC_if_dc() {
	local dc
	dc=`crmadmin -D 2>/dev/null | awk '{print $NF}'`
	if [ "$WE" = "$dc" ]; then
		touch $1/DC
	fi
}
corosync_blackbox() {
	local from_time=$1
	local to_time=$2
	local outf=$3
	local inpf
	inpf=`find_files /var/lib/corosync $from_time $to_time | grep -w fdata`
	if [ -f "$inpf" ]; then
		corosync-blackbox > $outf
		touch -r $inpf $outf
	fi
}
getconfigurations() {
	local conf
	local dest=$1
	for conf in $CONFIGURATIONS; do
		if [ -f $conf ]; then
			cp -p $conf $dest
		elif [ -d $conf ]; then
			(
			cd `dirname $conf` &&
			tar cf - `basename $conf` | (cd $dest && tar xf -)
			)
		fi
	done
}


#
# some basic system info and stats
#
sys_info() {
	cluster_info
	hb_report -V # our info
	echo "resource-agents: `grep 'Build version:' /usr/lib/ocf/lib/heartbeat/ocf-shellfuncs`"
	crm_info
	pkg_versions $PACKAGES
	skip_lvl 1 || verify_packages $PACKAGES
	echo "Platform: `uname`"
	echo "Kernel release: `uname -r`"
	echo "Architecture: `uname -m`"
	[ `uname` = Linux ] &&
		echo "Distribution: `distro`"
}
sys_stats() {
	set -x
	uname -n
	uptime
	ps axf
	ps auxw
	top -b -n 1
	ifconfig -a
	ip addr list
	netstat -i
	arp -an
	test -d /proc && {
		cat /proc/cpuinfo
	}
	lsscsi
	lspci
	mount
	# df can block, run in background, allow for 5 seconds (!)
	local maxcnt=5
	df &
	while kill -0 $! >/dev/null 2>&1; do
		sleep 1
		if [ $maxcnt -le 0 ]; then
			warning "df appears to be hanging, continuing without it"
			break
		fi
		maxcnt=$((maxcnt-1))
	done
	set +x
}
time_status() {
	date
	ntpdc -pn
}
dlm_dump() {
	if which dlm_tool >/dev/null 2>&1 ; then
		echo NOTICE - Lockspace overview:
		dlm_tool ls
		dlm_tool ls | grep name |
			while read X N ; do 
				echo NOTICE - Lockspace $N:
				dlm_tool lockdump $N
			done
		echo NOTICE - Lockspace history:
		dlm_tool dump
	fi
}


#
# replace sensitive info with '****'
#
sanitize() {
	local f rc
	for f in $1/$B_CONF; do
		[ -f "$f" ] && sanitize_one $f
	done
	rc=0
	for f in $1/$CIB_F $1/pengine/*; do
		if [ -f "$f" ]; then
			if [ "$DO_SANITIZE" ]; then
				sanitize_one $f
			else
				test_sensitive_one $f && rc=1
			fi
		fi
	done
	[ $rc -ne 0 ] && {
		warning "some PE or CIB files contain possibly sensitive data"
		warning "you may not want to send this report to a public mailing list"
	}
}

#
# remove duplicates if files are same, make links instead
#
consolidate() {
	for n in $NODES; do
		if [ -f $1/$2 ]; then
			rm $1/$n/$2
		else
			mv $1/$n/$2 $1
		fi
		ln -s ../$2 $1/$n
	done
}

#
# some basic analysis of the report
#
checkcrmvfy() {
	for n in $NODES; do
		if [ -s $1/$n/$CRM_VERIFY_F ]; then
			echo "WARN: crm_verify reported warnings at $n:"
			cat $1/$n/$CRM_VERIFY_F
		fi
	done
}
checkbacktraces() {
	for n in $NODES; do
		[ -s $1/$n/$BT_F ] && {
			echo "WARN: coredumps found at $n:"
			egrep 'Core was generated|Program terminated' \
					$1/$n/$BT_F |
				sed 's/^/	/'
		}
	done
}
checkpermissions() {
	for n in $NODES; do
		if [ -s $1/$n/$PERMISSIONS_F ]; then
			echo "WARN: problem with permissions/ownership at $n:"
			cat $1/$n/$PERMISSIONS_F
		fi
	done
}
checklogs() {
	local logs pattfile l n
	logs=$(find $1 -name $HALOG_F;
		for l in $EXTRA_LOGS; do find $1/* -name `basename $l`; done)
	[ "$logs" ] || return
	pattfile=`mktemp` ||
		fatal "cannot create temporary files"
	add_tmpfiles $pattfile
	for p in $LOG_PATTERNS; do
		echo "$p"
	done > $pattfile
	echo ""
	echo "Log patterns:"
	for n in $NODES; do
		cat $logs | grep -f $pattfile
	done
}

#
# check if files have same content in the cluster
#
cibdiff() {
	local d1 d2
	d1=`dirname $1`
	d2=`dirname $2`
	if [ -f $d1/RUNNING -a -f $d2/RUNNING ] ||
		[ -f $d1/STOPPED -a -f $d2/STOPPED ]; then
		if which crm_diff > /dev/null 2>&1; then
			crm_diff -c -n $1 -o $2
		else
			info "crm_diff(8) not found, cannot diff CIBs"
		fi
	else
		echo "can't compare cibs from running and stopped systems"
	fi
}
txtdiff() {
	diff -bBu $1 $2
}
diffcheck() {
	[ -f "$1" ] || {
		echo "$1 does not exist"
		return 1
	}
	[ -f "$2" ] || {
		echo "$2 does not exist"
		return 1
	}
	case `basename $1` in
	$CIB_F)
		cibdiff $1 $2;;
	$B_CONF)
		txtdiff $1 $2;; # confdiff?
	*)
		txtdiff $1 $2;;
	esac
}
analyze_one() {
	local rc node0 n
	rc=0
	node0=""
	for n in $NODES; do
		if [ "$node0" ]; then
			diffcheck $1/$node0/$2 $1/$n/$2
			rc=$(($rc+$?))
		else
			node0=$n
		fi
	done
	return $rc
}
analyze() {
	local f flist
	flist="$HOSTCACHE $MEMBERSHIP_F $CIB_F $CRM_MON_F $B_CONF logd.cf $SYSINFO_F"
	for f in $flist; do
		printf "Diff $f... "
		ls $1/*/$f >/dev/null 2>&1 || {
			echo "no $1/*/$f :/"
			continue
		}
		if analyze_one $1 $f; then
			echo "OK"
			[ "$f" != $CIB_F ] &&
				consolidate $1 $f
		else
			echo ""
		fi
	done
	checkcrmvfy $1
	checkbacktraces $1
	checkpermissions $1
	checklogs $1
}
events_all() {
	local Epatt title p
	Epatt=`echo "$EVENT_PATTERNS" |
		while read title p; do [ -n "$p" ] && echo -n "|$p"; done |
		sed 's/.//'
	`
	grep -E "$Epatt" $1
}
events() {
	local destdir n
	destdir=$1
	if [ -f $destdir/$HALOG_F ]; then
		events_all $destdir/$HALOG_F > $destdir/events.txt
		for n in $NODES; do
			awk "\$4==\"$n\"" $destdir/events.txt > $destdir/$n/events.txt
		done
	else
		for n in $NODES; do
			[ -f $destdir/$n/$HALOG_F ] ||
				continue
			events_all $destdir/$n/$HALOG_F > $destdir/$n/events.txt
		done
	fi
}

#
# description template, editing, and other notes
#
mktemplate() {
	cat<<EOF
Please edit this template and describe the issue/problem you
encountered. Then, post to
	Linux-HA@lists.linux-ha.org
or file a bug at
	http://developerbugs.linux-foundation.org/

See http://linux-ha.org/wiki/ReportingProblems for detailed
description on how to report problems.

Thank you.

Date: `date`
By: $PROG $userargs
Subject: [short problem description]
Severity: [choose one] enhancement minor normal major critical blocking
Component: [choose one] CRM LRM CCM RA fencing $CLUSTER_TYPE comm GUI tools other

Detailed description:
---
[...]
---

EOF

	if [ -f $WORKDIR/$SYSINFO_F ]; then
		echo "Common system info found:"
		cat $WORKDIR/$SYSINFO_F
	else
		for n in $NODES; do
			if [ -f $WORKDIR/$n/$SYSINFO_F ]; then
				echo "System info $n:"
				sed 's/^/	/' $WORKDIR/$n/$SYSINFO_F
			fi
		done
	fi
}
edittemplate() {
	local ec
	if ec=`pickfirst $EDITOR vim vi emacs nano`; then
		$ec $1
	else
		warning "could not find a text editor"
	fi
}
pickcompress() {
	if COMPRESS_PROG=`pickfirst bzip2 gzip xz`; then
		if [ "$COMPRESS_PROG" = xz ]; then
			COMPRESS_EXT=.xz
		elif [ "$COMPRESS_PROG" = bzip2 ]; then
			COMPRESS_EXT=.bz2
		else
			COMPRESS_EXT=.gz
		fi
	else
		warning "could not find a compression program; the resulting tarball may be huge"
		COMPRESS_PROG=cat
		COMPRESS_EXT=
	fi
}
# get the right part of the log
getlog() {
	local cnt
	local outf
	outf=$WORKDIR/$HALOG_F

	if [ "$HA_LOG" ]; then  # log provided by the user?
		[ -f "$HA_LOG" ] || {  # not present
			is_collector ||  # warning if not on slave
				warning "$HA_LOG not found; we will try to find log ourselves"
			HA_LOG=""
		}
	fi
	if [ "$HA_LOG" = "" ]; then
		HA_LOG=`findlog`
		[ "$HA_LOG" ] &&
			cnt=`fgrep -c $UNIQUE_MSG < $HA_LOG`
	fi
	if [ "$HA_LOG" = "" -o ! -f "$HA_LOG" ]; then
		if [ "$CTS" ]; then
			cts_findlogseg $CTS > $outf
		else
			warning "no log at $WE"
		fi
		return
	fi
	if [ "$cnt" ] && [ $cnt -gt 1 -a $cnt -eq $NODECNT ]; then
		MASTER_IS_HOSTLOG=1
		info "found the central log!"
	fi

	if [ "$NO_str2time" ]; then
		warning "a log found; but we cannot slice it"
		warning "please install the perl Date::Parse module"
	elif [ "$CTS" ]; then
		cts_findlogseg $CTS $HA_LOG > $outf
	else
		getstampproc=`find_getstampproc < $HA_LOG`
		if [ "$getstampproc" ]; then
			export getstampproc # used by linetime
			dumplogset $HA_LOG $FROM_TIME $TO_TIME > $outf &&
				loginfo $HA_LOG > $outf.info ||
				fatal "disk full"
		else
			warning "could not figure out the log format of $HA_LOG"
		fi
	fi
}
#
# get all other info (config, stats, etc)
#
collect_info() {
	local l
	sys_info > $WORKDIR/$SYSINFO_F 2>&1 &
	sys_stats > $WORKDIR/$SYSSTATS_F 2>&1 &
	getconfig $WORKDIR
	getpeinputs $FROM_TIME $TO_TIME $WORKDIR &
	crmconfig $WORKDIR &
	skip_lvl 1 || touch_DC_if_dc $WORKDIR &
	getbacktraces $FROM_TIME $TO_TIME $WORKDIR/$BT_F
	getconfigurations $WORKDIR
	check_perms > $WORKDIR/$PERMISSIONS_F 2>&1
	dlm_dump > $WORKDIR/$DLM_DUMP_F 2>&1
	time_status > $WORKDIR/$TIME_F 2>&1
	corosync_blackbox $FROM_TIME $TO_TIME $WORKDIR/$COROSYNC_RECORDER_F
	getratraces $FROM_TIME $TO_TIME $WORKDIR
	wait
	skip_lvl 1 || sanitize $WORKDIR

	for l in $EXTRA_LOGS; do
		[ "$NO_str2time" ] && break
		[ ! -f "$l" ] && continue
		if [ "$l" = "$HA_LOG" -a "$l" != "$HALOG_F" ]; then
			ln -s $HALOG_F $WORKDIR/`basename $l`
			continue
		fi
		getstampproc=`find_getstampproc < $l`
		if [ "$getstampproc" ]; then
			export getstampproc # used by linetime
			dumplogset $l $FROM_TIME $TO_TIME > $WORKDIR/`basename $l` &&
				loginfo $l > $WORKDIR/`basename $l`.info ||
				fatal "disk full"
		else
			warning "could not figure out the log format of $l"
		fi
	done
}
finalword() {
	if [ "$COMPRESS" = "1" ]; then
		echo "The report is saved in $DESTDIR/$DEST.tar$COMPRESS_EXT"
	else
		echo "The report is saved in $DESTDIR/$DEST"
	fi
	echo " "
	echo "Thank you for taking time to create this report."
}

[ $# -eq 0 ] && usage

# check for the major prereq for a) parameter parsing and b)
# parsing logs
#
NO_str2time=""
t=`str2time "12:00"`
if [ "$t" = "" ]; then
	NO_str2time=1
	is_collector ||
		fatal "please install the perl Date::Parse module"
fi

WE=`uname -n`  # who am i?
tmpdir=`mktemp -t -d .hb_report.workdir.XXXXXX` ||
	fatal "disk full"
add_tmpfiles $tmpdir
WORKDIR=$tmpdir

#
# part 1: get and check options; and the destination
#
if ! is_collector; then
	setvarsanddefaults
	userargs="$@"
	DESTDIR=.
	DEST="hb_report-"`date +"%a-%d-%b-%Y"`
	while getopts f:t:l:u:X:p:L:e:E:n:MSDCZAVsvhdQ o; do
		case "$o" in
			h) usage;;
			V) version;;
			f)
				if echo "$OPTARG" | grep -qs '^cts:'; then
					FROM_TIME=0 # to be calculated later
					CTS=`echo "$OPTARG" | sed 's/cts://'`
					DEST="cts-$CTS-"`date +"%a-%d-%b-%Y"`
				else
					FROM_TIME=`str2time "$OPTARG"`
					chktime "$FROM_TIME" "$OPTARG"
				fi
			;;
			t) TO_TIME=`str2time "$OPTARG"`
				chktime "$TO_TIME" "$OPTARG"
			;;
			n) NODES_SOURCE=user
				USER_NODES="$USER_NODES $OPTARG"
			;;
			u) SSH_USER="$OPTARG";;
			X) SSH_OPTS="$SSH_OPTS $OPTARG";;
			l) HA_LOG="$OPTARG";;
			e) EDITOR="$OPTARG";;
			p) SANITIZE="$SANITIZE $OPTARG";;
			s) DO_SANITIZE="1";;
			Q) SKIP_LVL=$((SKIP_LVL + 1));;
			L) LOG_PATTERNS="$LOG_PATTERNS $OPTARG";;
			S) NO_SSH=1;;
			D) NO_DESCRIPTION=1;;
			C) : ;;
			Z) FORCE_REMOVE_DEST=1;;
			M) EXTRA_LOGS="";;
			E) EXTRA_LOGS="$EXTRA_LOGS $OPTARG";;
			A) USER_CLUSTER_TYPE="openais";;
			v) VERBOSITY=$((VERBOSITY + 1));;
			d) COMPRESS="";;
			[?]) usage short;;
		esac
	done
	shift $(($OPTIND-1))
	[ $# -gt 1 ] && usage short
	set_dest $*
	[ "$FROM_TIME" ] || usage short
	WORKDIR=$tmpdir/$DEST
else
	WORKDIR=$tmpdir/$DEST/$WE
fi

mkdir -p $WORKDIR
[ -d $WORKDIR ] || no_dir

if is_collector; then
	cat > $WORKDIR/.env
	. $WORKDIR/.env
fi

[ $VERBOSITY -gt 1 ] && {
	is_collector || {
		info "high debug level, please read debug.out"
	}
	PS4='+ `date +"%T"`: ${FUNCNAME[0]:+${FUNCNAME[0]}:}${LINENO}: '
	if echo "$SHELL" | grep bash > /dev/null &&
			[ ${BASH_VERSINFO[0]} = "4" ]; then
		exec 3>>$WORKDIR/debug.out
		BASH_XTRACEFD=3
	else
		exec 2>>$WORKDIR/debug.out
	fi
	set -x
}

compatibility_pcmk

# allow user to enforce the cluster type
# if not, then it is found out on _all_ nodes
if [ -z "$USER_CLUSTER_TYPE" ]; then
	CLUSTER_TYPE=`get_cluster_type`
else
	CLUSTER_TYPE=$USER_CLUSTER_TYPE
fi

# the very first thing we must figure out is which cluster
# stack is used
CORES_DIRS="`2>/dev/null ls -d $HA_VARLIB/cores $PCMK_LIB/cores | uniq`"
PACKAGES="pacemaker libpacemaker3 
pacemaker-pygui pacemaker-pymgmt pymgmt-client
openais libopenais2 libopenais3 corosync libcorosync4
resource-agents cluster-glue libglue2 ldirectord
heartbeat heartbeat-common heartbeat-resources libheartbeat2
ocfs2-tools ocfs2-tools-o2cb ocfs2console
ocfs2-kmp-default ocfs2-kmp-pae ocfs2-kmp-xen ocfs2-kmp-debug ocfs2-kmp-trace
drbd drbd-kmp-xen drbd-kmp-pae drbd-kmp-default drbd-kmp-debug drbd-kmp-trace
drbd-heartbeat drbd-pacemaker drbd-utils drbd-bash-completion drbd-xen
lvm2 lvm2-clvm cmirrord
libdlm libdlm2 libdlm3
hawk ruby lighttpd
kernel-default kernel-pae kernel-xen
glibc
"
case "$CLUSTER_TYPE" in
openais)
	CONF=/etc/corosync/corosync.conf # corosync?
	if test -f $CONF; then
		CORES_DIRS="$CORES_DIRS /var/lib/corosync"
	else
		CONF=/etc/ais/openais.conf
		CORES_DIRS="$CORES_DIRS /var/lib/openais"
	fi
	CF_SUPPORT=$HA_NOARCHBIN/openais_conf_support.sh
	MEMBERSHIP_TOOL_OPTS=""
	unset HOSTCACHE HB_UUID_F
	;;
heartbeat)
	CONF=$HA_CF
	CF_SUPPORT=$HA_NOARCHBIN/ha_cf_support.sh
	MEMBERSHIP_TOOL_OPTS="-H"
	;;
esac
B_CONF=`basename $CONF`

if test -f "$CF_SUPPORT"; then
	. $CF_SUPPORT
else
	fatal "no stack specific support: $CF_SUPPORT"
fi

if [ "x$CTS" = "x" ] || is_collector; then
	getlogvars
	debug "log settings: facility=$HA_LOGFACILITY logfile=$HA_LOGFILE debugfile=$HA_DEBUGFILE"
elif ! is_collector; then
	ctslog=`findmsg "CTS: Stack:" | awk '{print $1}'`
	debug "Using CTS control file: $ctslog"
	USER_NODES=`grep CTS: $ctslog | grep -v debug: | grep " \* " | sed s:.*\\\*::g | sort -u  | tr '\\n' ' '`
	HA_LOGFACILITY=`findmsg "CTS:.*Environment.SyslogFacility" | awk '{print $NF}'`
	NODES_SOURCE=user
fi

# the goods
ANALYSIS_F=analysis.txt
DESCRIPTION_F=description.txt
HALOG_F=ha-log.txt
BT_F=backtraces.txt
SYSINFO_F=sysinfo.txt
SYSSTATS_F=sysstats.txt
DLM_DUMP_F=dlm_dump.txt
TIME_F=time.txt
export ANALYSIS_F DESCRIPTION_F HALOG_F BT_F SYSINFO_F SYSSTATS_F DLM_DUMP_F TIME_F
CRM_MON_F=crm_mon.txt
MEMBERSHIP_F=members.txt
HB_UUID_F=hb_uuid.txt
HOSTCACHE=hostcache
CRM_VERIFY_F=crm_verify.txt
PERMISSIONS_F=permissions.txt
CIB_F=cib.xml
CIB_TXT_F=cib.txt
COROSYNC_RECORDER_F=fdata.txt
export CRM_MON_F MEMBERSHIP_F CRM_VERIFY_F CIB_F CIB_TXT_F HB_UUID_F PERMISSIONS_F
export COROSYNC_RECORDER_F
CONFIGURATIONS="/etc/drbd.conf /etc/drbd.d /etc/booth/booth.conf"
export CONFIGURATIONS

THIS_IS_NODE=""
if ! is_collector; then
	MASTER_NODE=$WE
	NODES=`getnodes`
	debug "nodes: `echo $NODES`"
fi
NODECNT=`echo $NODES | wc -w`
if [ "$NODECNT" = 0 ]; then
	fatal "could not figure out a list of nodes; is this a cluster node?"
fi
if echo $NODES | grep -wqs $WE; then # are we a node?
	THIS_IS_NODE=1
fi

# this only on master
if ! is_collector; then

	# if this is not a node, then some things afterwards might
	# make no sense (not work)
	if ! is_node && [ "$NODES_SOURCE" != user ]; then
		warning "this is not a node and you didn't specify a list of nodes using -n"
	fi
#
# part 2: ssh business
#
	# find out if ssh works
	if [ -z "$NO_SSH" ]; then
		# if the ssh user was supplied, consider that it
		# works; helps reduce the number of ssh invocations
		findsshuser
		if [ -n "$SSH_USER" ]; then
			SSH_OPTS="$SSH_OPTS -o User=$SSH_USER"
		fi
	fi
	# assume that only root can collect data
	SUDO=""
	if [ -z "$SSH_USER" -a `id -u` != 0 ] ||
			[ -n "$SSH_USER" -a "$SSH_USER" != root ]; then
		debug "ssh user other than root, use sudo"
		SUDO="sudo -u root"
	fi
	LOCAL_SUDO=""
	if [ `id -u` != 0 ]; then
		debug "local user other than root, use sudo"
		LOCAL_SUDO="sudo -u root"
	fi
fi

if is_collector && [ "$HA_LOGFACILITY" ]; then
	logmark $HA_LOGFACILITY.$HA_LOGLEVEL $UNIQUE_MSG
	# allow for the log message to get (hopefully) written to the
	# log file
	sleep 1
fi

#
# part 4: find the logs and cut out the segment for the period
#

# if the master is also a node, getlog is going to be invoked
# from the collector
(is_master && is_node) ||
	getlog

if ! is_collector; then
	for node in $NODES; do
		if node_needs_pwd $node; then
			info "Please provide password for `say_ssh_user` at $node"
			info "Note that collecting data will take a while."
			start_slave_collector $node
		else
			start_slave_collector $node &
			SLAVEPIDS="$SLAVEPIDS $!"
		fi
	done
fi

#
# part 5: endgame:
#         slaves tar their results to stdout, the master waits
#         for them, analyses results, asks the user to edit the
#         problem description template, and prints final notes
#
if is_collector; then
	collect_info
	(cd $WORKDIR/.. && tar -h -cf - $WE)
else
	if [ -n "$SLAVEPIDS" ]; then
		wait $SLAVEPIDS
	fi
	analyze $WORKDIR > $WORKDIR/$ANALYSIS_F &
	events $WORKDIR &
	mktemplate > $WORKDIR/$DESCRIPTION_F
	[ "$NO_DESCRIPTION" ] || {
		echo press enter to edit the problem description...
		read junk
		edittemplate $WORKDIR/$DESCRIPTION_F
	}
	wait
	if [ "$COMPRESS" = "1" ]; then
		pickcompress
		(cd $WORKDIR/.. && tar cf - $DEST) | $COMPRESS_PROG > $DESTDIR/$DEST.tar$COMPRESS_EXT
	else
		mv $WORKDIR $DESTDIR
	fi
	finalword
fi