path: root/heartbeat/LVM

#!/bin/sh
#
# 
# LVM
#
# Description:	Manages an LVM volume as an HA resource
#
#
# Author:	Alan Robertson
# Support:	users@clusterlabs.org
# License:	GNU General Public License (GPL)
# Copyright:	(C) 2002 - 2005 International Business Machines, Inc.
#
#	This code significantly inspired by the LVM resource
#	in FailSafe by Lars Marowsky-Bree
#
#
# An example usage in /etc/ha.d/haresources: 
#			 node1	10.0.0.170 ServeRAID::1::1 LVM::myvolname
#
# See usage() function below for more details...
#
#		OCF parameters are as below:
#		OCF_RESKEY_volgrpname
#		
#######################################################################
# Initialization:

: ${OCF_FUNCTIONS_DIR=${OCF_ROOT}/lib/heartbeat}
. ${OCF_FUNCTIONS_DIR}/ocf-shellfuncs

# Parameter defaults

OCF_RESKEY_volgrpname_default=""
OCF_RESKEY_exclusive_default="false"
OCF_RESKEY_tag_default="pacemaker"
OCF_RESKEY_partial_activation_default="false"

: ${OCF_RESKEY_volgrpname=${OCF_RESKEY_volgrpname_default}}
: ${OCF_RESKEY_exclusive=${OCF_RESKEY_exclusive_default}}
: ${OCF_RESKEY_tag=${OCF_RESKEY_tag_default}}
: ${OCF_RESKEY_partial_activation=${OCF_RESKEY_partial_activation_default}}

#######################################################################


usage() {
	methods=`LVM_methods`
	methods=`echo $methods | tr ' ' '|'`
	cat <<EOF
	usage: $0 $methods

	$0 manages an	Linux Volume Manager volume (LVM) as an HA resource

	The 'start' operation brings the given volume online
	The 'stop' operation takes the given volume offline
	The 'status' operation reports whether the volume is available
	The 'monitor' operation reports whether the volume seems present
	The 'validate-all' operation checks whether the OCF parameters are valid
	The 'meta-data' operation show meta data 
	The 'methods' operation reports on the methods $0 supports

EOF
}

meta_data() {
	cat <<EOF
<?xml version="1.0"?>
<!DOCTYPE resource-agent SYSTEM "ra-api-1.dtd">
<resource-agent name="LVM" version="1.0">
<version>1.0</version>

<longdesc lang="en">
Resource script for LVM. It manages an Linux Volume Manager volume (LVM) 
as an HA resource. 
</longdesc>
<shortdesc lang="en">Controls the availability of an LVM Volume Group</shortdesc>

<parameters>
<parameter name="volgrpname" unique="1" required="1">
<longdesc lang="en">
The name of volume group.
</longdesc>
<shortdesc lang="en">Volume group name</shortdesc>
<content type="string" default="${OCF_RESKEY_volgrpname_default}" />
</parameter>
<parameter name="exclusive" unique="0" required="0">
<longdesc lang="en">
If set, the volume group will be activated exclusively.  This option works one of
two ways.  If the volume group has the cluster attribute set, then the volume group
will be activated exclusively using clvmd across the cluster.  If the cluster attribute
is not set, the volume group will be activated exclusively using a tag and the volume_list 
filter. When the tag option is in use, the volume_list in lvm.con must be initialized. This 
can be as simple as setting 'volume_list = []' depending on your setup.
</longdesc>
<shortdesc lang="en">Exclusive activation</shortdesc>
<content type="boolean" default="${OCF_RESKEY_exclusive_default}" />
</parameter>

<parameter name="tag" unique="0" required="0">
<longdesc lang="en">
If "exclusive" is set on a non clustered volume group, this overrides the tag to be used.
</longdesc>
<shortdesc lang="en">Exclusive activation tag</shortdesc>
<content type="string" default="${OCF_RESKEY_tag_default}" />
</parameter>

<parameter name="partial_activation" unique="0" required="0">
<longdesc lang="en">
If set, the volume group will be activated partially even with some
physical volumes missing. It helps to set to true when using mirrored
logical volumes.
</longdesc>
<shortdesc lang="en">Activate VG partially when missing PVs</shortdesc>
<content type="string" default="${OCF_RESKEY_partial_activation_default}" />
</parameter>

</parameters>

<actions>
<action name="start" timeout="30s" />
<action name="stop" timeout="30s" />
<action name="status" timeout="30s" />
<action name="monitor" depth="0" timeout="30s" interval="10s" />
<action name="methods" timeout="5s" />
<action name="meta-data" timeout="5s" />
<action name="validate-all" timeout="5s" />
</actions>
</resource-agent>
EOF
}

#
# methods: What methods/operations do we support?
#
LVM_methods() {
	cat <<EOF
	start
	stop
	status
	monitor
	methods
	validate-all
	meta-data
	usage
EOF
}

##
#
# plain = normal (non-exclusive) local activation
# tag = tagged-exclusive activation
# clvm = clvm-exclusive activation
#
# the mode specific implementation is in lvm-$mode.sh
##

set_lvm_mode() {
	local mode

	if ocf_is_true "$OCF_RESKEY_exclusive"; then
		case $(vgs -o attr --noheadings $OCF_RESKEY_volgrpname | tr -d ' ') in
		?????c*)
			mode="clvm" ;;
		*)
			mode="tag" ;;
		esac
	else
		mode="plain"
	fi

	. ${OCF_FUNCTIONS_DIR}/lvm-${mode}.sh
}

#
#	Return LVM status (silently)
#
LVM_status() {
	local rc=1
	loglevel="debug"

	# Set the log level of the error message
	if [ "X${2}" = "X" ]; then
		loglevel="err"
		if ocf_is_probe; then
			loglevel="warn"
		else 
			if [ ${OP_METHOD} = "stop" ]; then
				loglevel="info"
			fi
		fi
	fi
	
	if [ -d /dev/$1 ]; then
		test "`cd /dev/$1 && ls`" != ""
		rc=$?
		if [ $rc -ne 0 ]; then
			ocf_exit_reason "VG $1 with no logical volumes is not supported by this RA!"
		fi
	fi

	if [ $rc -ne 0 ]; then
		ocf_log $loglevel "LVM Volume $1 is not available (stopped)"
		rc=$OCF_NOT_RUNNING
	else
		lvm_status
		rc=$?
	fi

	if [ "X${2}" = "X" ]; then
		# status call return
		return $rc
	fi

	# Report on LVM volume status to stdout...
	if [ $rc -eq 0 ]; then
		echo "Volume $1 is available (running)"
	else
		echo "Volume $1 is not available (stopped)"
	fi
	return $rc
}

#
#	Enable LVM volume
#
LVM_start() {
	local vg=$1

	# systemd drop-in to stop process before storage services during
	# shutdown/reboot
	if systemd_is_running ; then
		systemd_drop_in "99-LVM" "After" "blk-availability.service"
	fi

	# TODO: This MUST run vgimport as well
	ocf_log info "Activating volume group $vg"
	if [ "$LVM_MAJOR" -eq "1" ]; then
		ocf_run vgscan $vg
	else
		ocf_run vgscan
	fi

	lvm_pre_activate || exit
	ocf_run vgchange $vgchange_activate_options $vg
	lvm_post_activate $?

	if LVM_status $vg; then
		: OK Volume $vg activated just fine!
		return $OCF_SUCCESS 
	else
		ocf_exit_reason "LVM: $vg did not activate correctly"
		return $OCF_ERR_GENERIC
	fi
}

#
#	Disable the LVM volume
#
LVM_stop() {
	local res=$OCF_ERR_GENERIC
	local vg=$1

	if ! vgs $vg > /dev/null 2>&1; then
		ocf_log info "Volume group $vg not found"
		return $OCF_SUCCESS
	fi

	ocf_log info "Deactivating volume group $vg"

	lvm_pre_deactivate || exit

	for i in $(seq 10)
	do
		ocf_run vgchange $vgchange_deactivate_options $vg
		res=$?
		if LVM_status $vg; then
			ocf_exit_reason "LVM: $vg did not stop correctly"
			res=1
		fi

		if [ $res -eq 0 ]; then
			break
		fi

		res=$OCF_ERR_GENERIC
		ocf_log warn "$vg still Active"
		ocf_log info "Retry deactivating volume group $vg"
		sleep 1
		which udevadm > /dev/null 2>&1 && udevadm settle --timeout=5
	done

	lvm_post_deactivate $res
}

#
#	Check whether the OCF instance parameters are valid
#
LVM_validate_all() {
	check_binary $AWK

	##
	# lvmetad is a daemon that caches lvm metadata to improve the
	# performance of LVM commands. This daemon should never be used when
	# volume groups exist that are being managed by the cluster. The lvmetad
	# daemon introduces a response lag, where certain LVM commands look like
	# they have completed (like vg activation) when in fact the command
	# is still in progress by the lvmetad.  This can cause reliability issues
	# when managing volume groups in the cluster.  For Example, if you have a
	# volume group that is a dependency for another application, it is possible
	# the cluster will think the volume group is activated and attempt to start
	# the application before volume group is really accesible... lvmetad is bad.
	##
	lvm dumpconfig global/use_lvmetad | grep 'use_lvmetad.*=.*1' > /dev/null 2>&1
	if [ $? -eq 0 ]; then
		# for now warn users that lvmetad is enabled and that they should disable it. In the
		# future we may want to consider refusing to start, or killing the lvmetad daemon.
		ocf_log warn "Disable lvmetad in lvm.conf. lvmetad should never be enabled in a clustered environment. Set use_lvmetad=0 and kill the lvmetad process"
	fi
	
	##
	# Off-the-shelf tests...
	##
	VGOUT=`vgck ${VOLUME} 2>&1`
	if [ $? -ne 0 ]; then
		# Inconsistency might be due to missing physical volumes, which doesn't 
		# automatically mean we should fail.  If partial_activation=true then 
		# we should let start try to handle it, or if no PVs are listed as
		# "unknown device" then another node may have marked a device missing 
		# where we have access to all of them and can start without issue. 
		if vgs -o pv_attr --noheadings $OCF_RESKEY_volgrpname 2>/dev/null | grep 'm' > /dev/null 2>&1; then
			case $(vgs -o attr --noheadings $OCF_RESKEY_volgrpname | tr -d ' ') in
			???p??*)
				if ! ocf_is_true "$OCF_RESKEY_partial_activation" ; then
					# We are missing devices and cannot activate partially
					ocf_exit_reason "Volume group [$VOLUME] has devices missing.  Consider partial_activation=true to attempt to activate partially"
					exit $OCF_ERR_GENERIC
				else
					# We are missing devices but are allowed to activate partially. 
					# Assume that caused the vgck failure and carry on
					ocf_log warn "Volume group inconsistency detected with missing device(s) and partial_activation enabled.  Proceeding with requested action."
				fi
				;;
			esac
			# else the vg is partial but all devices are accounted for, so another 
			# node must have marked the device missing.  Proceed.
		else
			# vgck failure was for something other than missing devices
			ocf_exit_reason "Volume group [$VOLUME] does not exist or contains error! ${VGOUT}"
			exit $OCF_ERR_GENERIC
		fi
	fi

	##
	# Does the Volume Group exist?
	##
	if [ "$LVM_MAJOR" = "1" ]; then
		VGOUT=`vgdisplay ${VOLUME} 2>&1`
	else
		VGOUT=`vgdisplay -v ${VOLUME} 2>&1`
	fi
	if [ $? -ne 0 ]; then
		ocf_exit_reason "Volume group [$VOLUME] does not exist or contains error! ${VGOUT}"
		exit $OCF_ERR_GENERIC
	fi

	if lvs --noheadings -o segtype | grep -q "cache"; then
		if ! lvs --noheadings -o cache_mode "$OCF_RESKEY_volgrpname" | grep -q "writethrough"; then
			ocf_log warn "LVM CACHE IS NOT IN WRITETHROUGH MODE. THIS IS NOT A SUPPORTED CONFIGURATION."
		fi
	fi

	if ocf_is_clone && ocf_is_true "$OCF_RESKEY_exclusive"; then
		ocf_exit_reason "cloned lvm resources can not be activated exclusively"
		exit $OCF_ERR_CONFIGURED
	fi

	lvm_validate_all
}

#
#	'main' starts here...
#

if
	[ $# -ne 1 ]
then
	usage
	exit $OCF_ERR_ARGS 
fi

case $1 in
	meta-data)	meta_data
		exit $OCF_SUCCESS;;

	methods)	LVM_methods
		exit $?;;

	usage)	usage
		exit $OCF_SUCCESS;;
	*)		;;
esac

if 
	[ -z "$OCF_RESKEY_volgrpname" ]
then
	ocf_exit_reason "You must identify the volume group name!"
	exit $OCF_ERR_CONFIGURED 
fi

# Get the LVM version number, for this to work we assume(thanks to panjiam):
# 
# LVM1 outputs like this
#
#	# vgchange --version
#	vgchange: Logical Volume Manager 1.0.3
#	Heinz Mauelshagen, Sistina Software	19/02/2002 (IOP 10)
#
# LVM2 and higher versions output in this format
#
#	# vgchange --version
#	LVM version:		 2.00.15 (2004-04-19)
#	Library version: 1.00.09-ioctl (2004-03-31)
#	Driver version:	4.1.0

LVM_VERSION=`vgchange --version 2>&1 | \
	$AWK '/Logical Volume Manager/ {print $5"\n"; exit; }
			 /LVM version:/ {printf $3"\n"; exit;}'`
rc=$?

if
	( [ $rc -ne 0 ] || [ -z "$LVM_VERSION" ] )
then
	ocf_exit_reason "LVM: $1 could not determine LVM version. Try 'vgchange --version' manually and modify $0 ?"
	exit $OCF_ERR_INSTALLED
fi
LVM_MAJOR="${LVM_VERSION%%.*}"

VOLUME=$OCF_RESKEY_volgrpname
OP_METHOD=$1

set_lvm_mode
lvm_init
if ocf_is_true "$OCF_RESKEY_partial_activation" ; then
	vgchange_activate_options="${vgchange_activate_options} --partial"
fi

# What kind of method was invoked?
case "$1" in

	start)
		LVM_validate_all
		LVM_start $VOLUME
		exit $?;;

	stop)	LVM_stop $VOLUME
		exit $?;;

	status)	LVM_status $VOLUME $1
		exit $?;;

	monitor)	LVM_status $VOLUME
		exit $?;;

	validate-all)	LVM_validate_all
		;;

	*)		usage
		exit $OCF_ERR_UNIMPLEMENTED;;
esac