common-ha: continuous grace_mon log messages in /var/log/messages

messages are seen on RHEL6.x and RHEL7.1 and earlier versions of
pacemaker. (And RHEL7.2 with RHEL7.1 pacemaker packages.)

It's not possible to query attrd attributes in the older version,
only set/update/clear them. The messages come from invalid attempts
to query the attributes.

However it is possible to query crm attributes. The fix here is to
create a "shadow" crm attribute for the attrd attribute. Changes are
made to both, queries are made on the crm attribute.

(Resource Agents "follow" the attrd attribute using constraint locations,
so we must keep the attrd attribute.)

Change-Id: I84ac1a80673e528d98b67b7d5062e21dcf744d4a
BUG: 1324509
Signed-off-by: Kaleb S KEITHLEY <kkeithle@redhat.com>
Reviewed-on: http://review.gluster.org/13919
Reviewed-by: Niels de Vos <ndevos@redhat.com>
Smoke: Gluster Build System <jenkins@build.gluster.com>
NetBSD-regression: NetBSD Build System <jenkins@build.gluster.org>
CentOS-regression: Gluster Build System <jenkins@build.gluster.com>
Tested-by: NetBSD Build System <jenkins@build.gluster.org>
Reviewed-by: soumya k <skoduri@redhat.com>

4 files changed, 48 insertions, 33 deletions

diff --git a/extras/ganesha/ocf/ganesha_grace b/extras/ganesha/ocf/ganesha_grace
index a82c9af417a..7c629f5e773 100644
--- a/extras/ganesha/ocf/ganesha_grace
+++ b/extras/ganesha/ocf/ganesha_grace
@@ -30,17 +30,17 @@
 . ${OCF_FUNCTIONS_DIR}/ocf-shellfuncs
 
 if [ -n "$OCF_DEBUG_LIBRARY" ]; then
-    . $OCF_DEBUG_LIBRARY
+	. $OCF_DEBUG_LIBRARY
 else
-    : ${OCF_FUNCTIONS_DIR=${OCF_ROOT}/lib/heartbeat}
-. ${OCF_FUNCTIONS_DIR}/ocf-shellfuncs
+	: ${OCF_FUNCTIONS_DIR=${OCF_ROOT}/lib/heartbeat}
+	. ${OCF_FUNCTIONS_DIR}/ocf-shellfuncs
 fi
 
 OCF_RESKEY_grace_active_default="grace-active"
 : ${OCF_RESKEY_grace_active=${OCF_RESKEY_grace_active_default}}
 
 ganesha_meta_data() {
-        cat <<END
+	cat <<END
 <?xml version="1.0"?>
 <!DOCTYPE resource-agent SYSTEM "ra-api-1.dtd">
 <resource-agent name="ganesha_grace">
@@ -93,9 +93,11 @@ esac
 
 ganesha_grace_start()
 {
-        rc=${OCF_ERR_GENERIC}
+	local rc=${OCF_ERR_GENERIC}
+	local short_host=$(hostname -s)
+
 	ocf_log debug "ganesha_grace_start()"
-	attr=$(attrd_updater -Q -n ${OCF_RESKEY_grace_active})
+	attr=$(crm_attribute --query --node=${short_host} --name=${OCF_RESKEY_grace_active})
 
 	# Three possibilities:
 	# 1. There is no attribute at all and attr_updater returns
@@ -116,7 +118,7 @@ ganesha_grace_start()
 	fi
 
 	# case 2
-	if [[ "${attr}" = *"value=\"1\"" ]]; then
+	if [[ "${attr}" = *"value=1" ]]; then
 		return ${OCF_SUCCESS}
 	fi
 
@@ -149,10 +151,12 @@ ganesha_grace_notify()
 
 ganesha_grace_monitor()
 {
-        rc=${OCF_ERR_GENERIC}
+	local rc=${OCF_ERR_GENERIC}
+	local short_host=$(hostname -s)
+
 	ocf_log debug "monitor"
 
-	attr=$(attrd_updater -Q -n ${OCF_RESKEY_grace_active})
+	attr=$(crm_attribute --query --node=${short_host} --name=${OCF_RESKEY_grace_active})
 
 	# if there is no attribute (yet), maybe it's because
 	# this RA started before ganesha_mon (nfs-mon) has had
@@ -162,7 +166,7 @@ ganesha_grace_monitor()
 		return ${OCF_SUCCESS}
 	fi
 
-	if [[ "${attr}" = *"value=\"1\"" ]]; then
+	if [[ "${attr}" = *"value=1" ]]; then
 		rc=${OCF_SUCCESS}
 	fi
 
@@ -187,8 +191,8 @@ status|monitor) ganesha_grace_monitor
 notify)         ganesha_grace_notify
 		;;
 *)              ganesha_grace_usage
-                exit ${OCF_ERR_UNIMPLEMENTED}
-                ;;
+		exit ${OCF_ERR_UNIMPLEMENTED}
+		;;
 esac
 
 rc=$?
diff --git a/extras/ganesha/ocf/ganesha_mon b/extras/ganesha/ocf/ganesha_mon
index f55cf7f2af3..65450b82209 100644
--- a/extras/ganesha/ocf/ganesha_mon
+++ b/extras/ganesha/ocf/ganesha_mon
@@ -30,10 +30,10 @@
 . ${OCF_FUNCTIONS_DIR}/ocf-shellfuncs
 
 if [ -n "${OCF_DEBUG_LIBRARY}" ]; then
-    . ${OCF_DEBUG_LIBRARY}
+	. ${OCF_DEBUG_LIBRARY}
 else
-    : ${OCF_FUNCTIONS_DIR=${OCF_ROOT}/lib/heartbeat}
-. ${OCF_FUNCTIONS_DIR}/ocf-shellfuncs
+	: ${OCF_FUNCTIONS_DIR=${OCF_ROOT}/lib/heartbeat}
+	. ${OCF_FUNCTIONS_DIR}/ocf-shellfuncs
 fi
 
 # Defaults
@@ -46,7 +46,7 @@ OCF_RESKEY_grace_delay_default="5"
 : ${OCF_RESKEY_grace_delay=${OCF_RESKEY_grace_delay_default}}
 
 ganesha_meta_data() {
-        cat <<END
+	cat <<END
 <?xml version="1.0"?>
 <!DOCTYPE resource-agent SYSTEM "ra-api-1.dtd">
 <resource-agent name="ganesha_mon">
@@ -113,13 +113,13 @@ ganesha_mon_start()
 {
 	ocf_log debug "ganesha_mon_start"
 	ganesha_mon_monitor
-        return $OCF_SUCCESS
+	return $OCF_SUCCESS
 }
 
 ganesha_mon_stop()
 {
 	ocf_log debug "ganesha_mon_stop"
-        return $OCF_SUCCESS
+	return $OCF_SUCCESS
 }
 
 ganesha_mon_monitor()
@@ -142,11 +142,22 @@ ganesha_mon_monitor()
 			ocf_log info "warning: attrd_updater -n ${OCF_RESKEY_ganesha_active} -v 1 failed"
 		fi
 
+		# ganesha_grace (nfs-grace) RA follows grace-active attr
+		# w/ constraint location
 		attrd_updater -n ${OCF_RESKEY_grace_active} -v 1
 		if [ $? -ne 0 ]; then
 			ocf_log info "warning: attrd_updater -n ${OCF_RESKEY_grace_active} -v 1 failed"
 		fi
 
+		# ganesha_mon (nfs-mon) and ganesha_grace (nfs-grace)
+		# track grace-active crm_attr (attr != crm_attr)
+		# we can't just use the attr as there's no way to query
+		# its value in RHEL6 pacemaker
+		crm_attribute --node=${short_host} --lifetime=forever --name=${OCF_RESKEY_grace_active} --update=1
+		if [ $? -ne 0 ]; then
+			ocf_log info "warning: crm_attribute --node=${short_host} --lifetime=forever --name=${OCF_RESKEY_grace_active} --update=1 failed"
+		fi
+
 		return ${OCF_SUCCESS}
 	fi
 
@@ -155,23 +166,23 @@ ganesha_mon_monitor()
 	#
 	# Meanwhile the ganesha_grace notify() runs when its
 	# nfs-grace resource is disabled on a node; which
-	# is triggered by clearing the ganesha-grace node
-	# attribute on this node.
+	# is triggered by clearing the grace-active attribute
+	# on this node.
 	#
 	# We need to allow time for it to run and put
 	# the remaining ganesha.nfsds into grace before
 	# initiating the VIP fail-over.
 
+	sleep ${OCF_RESKEY_grace_delay}
+
 	attrd_updater -D -n ${OCF_RESKEY_grace_active}
 	if [ $? -ne 0 ]; then
 		ocf_log info "warning: attrd_updater -D -n ${OCF_RESKEY_grace_active} failed"
 	fi
 
-	sleep ${OCF_RESKEY_grace_delay}
-
-	attrd_updater -D -n ${OCF_RESKEY_ganesha_active}
+	crm_attribute --node=${short_host} --name=${OCF_RESKEY_grace_active} --update=0
 	if [ $? -ne 0 ]; then
-		ocf_log info "warning: attrd_updater -D -n ${OCF_RESKEY_ganesha_active} failed"
+		ocf_log info "warning: crm_attribute --node=${short_host} --name=${OCF_RESKEY_grace_active} --update=0 failed"
 	fi
 
 	return ${OCF_SUCCESS}
@@ -193,8 +204,8 @@ stop)           ganesha_mon_stop
 status|monitor) ganesha_mon_monitor
 		;;
 *)              ganesha_mon_usage
-                exit ${OCF_ERR_UNIMPLEMENTED}
-                ;;
+		exit ${OCF_ERR_UNIMPLEMENTED}
+		;;
 esac
 
 rc=$?
diff --git a/extras/ganesha/ocf/ganesha_nfsd b/extras/ganesha/ocf/ganesha_nfsd
index a9d3e4d860f..29e333ca903 100644
--- a/extras/ganesha/ocf/ganesha_nfsd
+++ b/extras/ganesha/ocf/ganesha_nfsd
@@ -30,17 +30,17 @@
 . ${OCF_FUNCTIONS_DIR}/ocf-shellfuncs
 
 if [ -n "${OCF_DEBUG_LIBRARY}" ]; then
-    . ${OCF_DEBUG_LIBRARY}
+	. ${OCF_DEBUG_LIBRARY}
 else
-    : ${OCF_FUNCTIONS_DIR=${OCF_ROOT}/lib/heartbeat}
-. ${OCF_FUNCTIONS_DIR}/ocf-shellfuncs
+	: ${OCF_FUNCTIONS_DIR=${OCF_ROOT}/lib/heartbeat}
+	. ${OCF_FUNCTIONS_DIR}/ocf-shellfuncs
 fi
 
 OCF_RESKEY_ha_vol_mnt_default="/var/run/gluster/shared_storage"
 : ${OCF_RESKEY_ha_vol_mnt=${OCF_RESKEY_ha_vol_mnt_default}}
 
 ganesha_meta_data() {
-        cat <<END
+	cat <<END
 <?xml version="1.0"?>
 <!DOCTYPE resource-agent SYSTEM "ra-api-1.dtd">
 <resource-agent name="ganesha_nfsd">
@@ -156,8 +156,8 @@ stop)           ganesha_nfsd_stop
 status|monitor) ganesha_nfsd_monitor
 		;;
 *)              ganesha_nfsd_usage
-                exit ${OCF_ERR_UNIMPLEMENTED}
-                ;;
+		exit ${OCF_ERR_UNIMPLEMENTED}
+		;;
 esac
 
 rc=$?
diff --git a/extras/ganesha/scripts/ganesha-ha.sh b/extras/ganesha/scripts/ganesha-ha.sh
index d866220912f..7f1ea4c2da5 100644
--- a/extras/ganesha/scripts/ganesha-ha.sh
+++ b/extras/ganesha/scripts/ganesha-ha.sh
@@ -376,7 +376,7 @@ teardown_cluster()
         if [[ ${HA_CLUSTER_NODES} != *${server}* ]]; then
             logger "info: ${server} is not in config, removing"
 
-            pcs cluster stop ${server}
+            pcs cluster stop ${server} --force
             if [ $? -ne 0 ]; then
                 logger "warning: pcs cluster stop ${server} failed"
             fi