summaryrefslogtreecommitdiffstats
path: root/extras/ganesha
diff options
context:
space:
mode:
Diffstat (limited to 'extras/ganesha')
-rw-r--r--extras/ganesha/config/ganesha-ha.conf.sample7
-rw-r--r--extras/ganesha/ocf/Makefile.am1
-rw-r--r--extras/ganesha/ocf/ganesha_grace175
-rw-r--r--extras/ganesha/ocf/ganesha_mon153
-rw-r--r--extras/ganesha/ocf/ganesha_nfsd98
-rw-r--r--extras/ganesha/scripts/Makefile.am6
-rwxr-xr-xextras/ganesha/scripts/create-export-ganesha.sh79
-rwxr-xr-xextras/ganesha/scripts/dbus-send.sh117
-rw-r--r--extras/ganesha/scripts/ganesha-ha.sh849
-rwxr-xr-xextras/ganesha/scripts/generate-epoch.py48
10 files changed, 919 insertions, 614 deletions
diff --git a/extras/ganesha/config/ganesha-ha.conf.sample b/extras/ganesha/config/ganesha-ha.conf.sample
index 24054abfdff..c22892bde56 100644
--- a/extras/ganesha/config/ganesha-ha.conf.sample
+++ b/extras/ganesha/config/ganesha-ha.conf.sample
@@ -2,9 +2,6 @@
# must be unique within the subnet
HA_NAME="ganesha-ha-360"
#
-# The gluster server from which to mount the shared data volume.
-HA_VOL_SERVER="server1"
-#
# N.B. you may use short names or long names; you may not use IP addrs.
# Once you select one, stay with it as it will be mildly unpleasant to
# clean up if you switch later on. Ensure that all names - short and/or
@@ -18,5 +15,5 @@ HA_CLUSTER_NODES="server1,server2,..."
# Virtual IPs for each of the nodes specified above.
VIP_server1="10.0.2.1"
VIP_server2="10.0.2.2"
-#VIP_server1.lab.redhat.com="10.0.2.1"
-#VIP_server2.lab.redhat.com="10.0.2.2"
+#VIP_server1_lab_redhat_com="10.0.2.1"
+#VIP_server2_lab_redhat_com="10.0.2.2"
diff --git a/extras/ganesha/ocf/Makefile.am b/extras/ganesha/ocf/Makefile.am
index 6aed9548a0f..990a609f254 100644
--- a/extras/ganesha/ocf/Makefile.am
+++ b/extras/ganesha/ocf/Makefile.am
@@ -9,4 +9,3 @@ ocfdir = $(prefix)/lib/ocf
radir = $(ocfdir)/resource.d/heartbeat
ra_SCRIPTS = ganesha_grace ganesha_mon ganesha_nfsd
-
diff --git a/extras/ganesha/ocf/ganesha_grace b/extras/ganesha/ocf/ganesha_grace
index 75ec16c0fd1..825f7164597 100644
--- a/extras/ganesha/ocf/ganesha_grace
+++ b/extras/ganesha/ocf/ganesha_grace
@@ -30,14 +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">
@@ -51,19 +54,25 @@ resource agent for nfs-ganesha.
<shortdesc lang="en">Manages the user-space nfs-ganesha NFS server</shortdesc>
<parameters>
+<parameter name="grace_active">
+<longdesc lang="en">NFS-Ganesha grace active attribute</longdesc>
+<shortdesc lang="en">NFS-Ganesha grace active attribute</shortdesc>
+<content type="string" default="grace-active" />
+</parameter>
</parameters>
<actions>
<action name="start" timeout="40s" />
<action name="stop" timeout="40s" />
-<action name="status" depth="0" timeout="20s" interval="5s" />
-<action name="monitor" depth="0" timeout="20s" interval="5s" />
+<action name="status" timeout="20s" interval="60s" />
+<action name="monitor" depth="0" timeout="10s" interval="5s" />
+<action name="notify" timeout="10s" />
<action name="meta-data" timeout="20s" />
</actions>
</resource-agent>
END
-return $OCF_SUCCESS
+return ${OCF_SUCCESS}
}
ganesha_grace_usage() {
@@ -73,10 +82,10 @@ ganesha_grace_usage() {
# Make sure meta-data and usage always succeed
case $__OCF_ACTION in
meta-data) ganesha_meta_data
- exit $OCF_SUCCESS
+ exit ${OCF_SUCCESS}
;;
usage|help) ganesha_usage
- exit $OCF_SUCCESS
+ exit ${OCF_SUCCESS}
;;
*)
;;
@@ -84,81 +93,112 @@ esac
ganesha_grace_start()
{
- local result=""
- local resourcename=""
- local deadserver=""
- local tmpIFS=${IFS}
- local pid_file="/var/run/ganesha.nfsd.pid"
-
- # RHEL6 /etc/init.d/nfs-ganesha adds "-p /var/run/ganesha.nfsd.pid"
- # RHEL7 systemd does not. Would be nicer if all distros used the
- # same pid file.
- if [ -e /usr/lib/systemd/system/nfs-ganesha.service ]; then
- pid_file="/var/run/ganesha.pid"
+ local rc=${OCF_ERR_GENERIC}
+ local host=$(hostname -s)
+
+ ocf_log debug "ganesha_grace_start()"
+ # give ganesha_mon RA a chance to set the crm_attr first
+ # I mislike the sleep, but it's not clear that looping
+ # with a small sleep is necessarily better
+ # start has a 40sec timeout, so a 5sec sleep here is okay
+ sleep 5
+ attr=$(crm_attribute --query --node=${host} --name=${OCF_RESKEY_grace_active} 2> /dev/null)
+ if [ $? -ne 0 ]; then
+ host=$(hostname)
+ attr=$(crm_attribute --query --node=${host} --name=${OCF_RESKEY_grace_active} 2> /dev/null )
+ if [ $? -ne 0 ]; then
+ ocf_log info "grace start: crm_attribute --query --node=${host} --name=${OCF_RESKEY_grace_active} failed"
+ fi
+ fi
+
+ # Three possibilities:
+ # 1. There is no attribute at all and attr_updater returns
+ # a zero length string. This happens when
+ # ganesha_mon::monitor hasn't run at least once to set
+ # the attribute. The assumption here is that the system
+ # is coming up. We pretend, for now, that the node is
+ # healthy, to allow the system to continue coming up.
+ # It will cure itself in a few seconds
+ # 2. There is an attribute, and it has the value "1"; this
+ # node is healthy.
+ # 3. There is an attribute, but it has no value or the value
+ # "0"; this node is not healthy.
+
+ # case 1
+ if [[ -z "${attr}" ]]; then
+ return ${OCF_SUCCESS}
fi
- # logger "ganesha_grace_start()"
- # we're here because somewhere in the cluster one or more
- # of the ganesha.nfsds have died, triggering a floating IP
- # address to move. Resource constraint location rules ensure
- # that this is invoked before the floating IP is moved.
- if [ -e ${pid_file} -a \
- -d /proc/$(cat ${pid_file} ) ]; then
- # my ganesha.nfsd is still running
- # find out which one died?
-
- pcs status | grep dead_ip-1 | sort > /tmp/.pcs_status
-
- result=$(diff /var/run/ganesha/pcs_status /tmp/.pcs_status | grep '^>')
- if [[ ${result} ]]; then
- # logger "ganesha_grace_start(), ${result}"
- IFS=$'\n'
- for line in ${result}; do
- resourcename=$(echo ${line} | cut -f 1 | cut -d ' ' -f 3)
- deadserver=${resourcename%"-dead_ip-1"}
-
- if [[ ${deadserver} ]]; then
- # logger "ganesha_grace_start(), ${line}"
- # logger "ganesha_grace_start(), dbus-send --print-reply --system --dest=org.ganesha.nfsd /org/ganesha/nfsd/admin org.ganesha.nfsd.admin.grace string:${deadserver}"
- dbus-send --print-reply --system --dest=org.ganesha.nfsd /org/ganesha/nfsd/admin org.ganesha.nfsd.admin.grace string:${deadserver}
- if [ $? -ne 0 ]; then
- logger "warning: dbus-send --print-reply --system --dest=org.ganesha.nfsd /org/ganesha/nfsd/admin org.ganesha.nfsd.admin.grace string:${deadserver} failed"
- fi
- fi
- done
- IFS=${tmpIFS}
- fi
-
+ # case 2
+ if [[ "${attr}" = *"value=1" ]]; then
+ return ${OCF_SUCCESS}
fi
- return $OCF_SUCCESS
+
+ # case 3
+ return ${OCF_NOT_RUNNING}
}
ganesha_grace_stop()
{
- # logger "ganesha_grace_stop()"
- return $OCF_SUCCESS
+ ocf_log debug "ganesha_grace_stop()"
+ return ${OCF_SUCCESS}
+}
+
+ganesha_grace_notify()
+{
+ # since this is a clone RA we should only ever see pre-start
+ # or post-stop
+ mode="${OCF_RESKEY_CRM_meta_notify_type}-${OCF_RESKEY_CRM_meta_notify_operation}"
+ case "${mode}" in
+ pre-start | post-stop)
+ dbus-send --print-reply --system --dest=org.ganesha.nfsd /org/ganesha/nfsd/admin org.ganesha.nfsd.admin.grace string:${OCF_RESKEY_CRM_meta_notify_stop_uname}
+ if [ $? -ne 0 ]; then
+ ocf_log info "dbus-send --print-reply --system --dest=org.ganesha.nfsd /org/ganesha/nfsd/admin org.ganesha.nfsd.admin.grace string:${OCF_RESKEY_CRM_meta_notify_stop_uname} failed"
+ fi
+ ;;
+ esac
+
+ return ${OCF_SUCCESS}
}
ganesha_grace_monitor()
{
- # logger "ganesha_grace_monitor()"
- if [ ! -d /var/run/ganesha ]; then
- mkdir -p /var/run/ganesha
+ local host=$(hostname -s)
+
+ ocf_log debug "monitor"
+
+ attr=$(crm_attribute --query --node=${host} --name=${OCF_RESKEY_grace_active} 2> /dev/null)
+ if [ $? -ne 0 ]; then
+ host=$(hostname)
+ attr=$(crm_attribute --query --node=${host} --name=${OCF_RESKEY_grace_active} 2> /dev/null)
+ if [ $? -ne 0 ]; then
+ ocf_log info "crm_attribute --query --node=${host} --name=${OCF_RESKEY_grace_active} failed"
+ fi
+ fi
+
+ # if there is no attribute (yet), maybe it's because
+ # this RA started before ganesha_mon (nfs-mon) has had
+ # chance to create it. In which case we'll pretend
+ # everything is okay this time around
+ if [[ -z "${attr}" ]]; then
+ return ${OCF_SUCCESS}
fi
- pcs status | grep dead_ip-1 | sort > /var/run/ganesha/pcs_status
- return $OCF_SUCCESS
+
+ if [[ "${attr}" = *"value=1" ]]; then
+ return ${OCF_SUCCESS}
+ fi
+
+ return ${OCF_NOT_RUNNING}
}
ganesha_grace_validate()
{
- return $OCF_SUCCESS
+ return ${OCF_SUCCESS}
}
ganesha_grace_validate
-# logger "ganesha_grace ${OCF_RESOURCE_INSTANCE} $__OCF_ACTION"
-
# Translate each action into the appropriate function call
case $__OCF_ACTION in
start) ganesha_grace_start
@@ -167,14 +207,15 @@ stop) ganesha_grace_stop
;;
status|monitor) ganesha_grace_monitor
;;
+notify) ganesha_grace_notify
+ ;;
*) ganesha_grace_usage
- exit $OCF_ERR_UNIMPLEMENTED
- ;;
+ exit ${OCF_ERR_UNIMPLEMENTED}
+ ;;
esac
rc=$?
# The resource agent may optionally log a debug message
-ocf_log debug "${OCF_RESOURCE_INSTANCE} $__OCF_ACTION returned $rc"
+ocf_log debug "${OCF_RESOURCE_INSTANCE} ${__OCF_ACTION} returned $rc"
exit $rc
-
diff --git a/extras/ganesha/ocf/ganesha_mon b/extras/ganesha/ocf/ganesha_mon
index c8e7de9c45e..2b4a9d6da84 100644
--- a/extras/ganesha/ocf/ganesha_mon
+++ b/extras/ganesha/ocf/ganesha_mon
@@ -29,17 +29,24 @@
: ${OCF_FUNCTIONS_DIR=${OCF_ROOT}/lib/heartbeat}
. ${OCF_FUNCTIONS_DIR}/ocf-shellfuncs
-if [ -n "$OCF_DEBUG_LIBRARY" ]; then
- . $OCF_DEBUG_LIBRARY
+if [ -n "${OCF_DEBUG_LIBRARY}" ]; then
+ . ${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
-GRACE_DELAY=7
+# Defaults
+OCF_RESKEY_ganesha_active_default="ganesha-active"
+OCF_RESKEY_grace_active_default="grace-active"
+OCF_RESKEY_grace_delay_default="5"
+
+: ${OCF_RESKEY_ganesha_active=${OCF_RESKEY_ganesha_active_default}}
+: ${OCF_RESKEY_grace_active=${OCF_RESKEY_grace_active_default}}
+: ${OCF_RESKEY_grace_delay=${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">
@@ -53,19 +60,37 @@ resource agent for nfs-ganesha.
<shortdesc lang="en">Manages the user-space nfs-ganesha NFS server</shortdesc>
<parameters>
+<parameter name="ganesha_active">
+<longdesc lang="en">NFS-Ganesha daemon active attribute</longdesc>
+<shortdesc lang="en">NFS-Ganesha daemon active attribute</shortdesc>
+<content type="string" default="ganesha-active" />
+</parameter>
+<parameter name="grace_active">
+<longdesc lang="en">NFS-Ganesha grace active attribute</longdesc>
+<shortdesc lang="en">NFS-Ganesha grace active attribute</shortdesc>
+<content type="string" default="grace-active" />
+</parameter>
+<parameter name="grace_delay">
+<longdesc lang="en">
+NFS-Ganesha grace delay.
+When changing this, adjust the ganesha_grace RA's monitor interval to match.
+</longdesc>
+<shortdesc lang="en">NFS-Ganesha grace delay</shortdesc>
+<content type="string" default="5" />
+</parameter>
</parameters>
<actions>
<action name="start" timeout="40s" />
<action name="stop" timeout="40s" />
-<action name="status" depth="0" timeout="20s" interval="10s" />
+<action name="status" timeout="20s" interval="60s" />
<action name="monitor" depth="0" timeout="10s" interval="10s" />
<action name="meta-data" timeout="20s" />
</actions>
</resource-agent>
END
-return $OCF_SUCCESS
+return ${OCF_SUCCESS}
}
ganesha_mon_usage() {
@@ -73,12 +98,12 @@ ganesha_mon_usage() {
}
# Make sure meta-data and usage always succeed
-case $__OCF_ACTION in
+case ${__OCF_ACTION} in
meta-data) ganesha_meta_data
- exit $OCF_SUCCESS
+ exit ${OCF_SUCCESS}
;;
usage|help) ganesha_usage
- exit $OCF_SUCCESS
+ exit ${OCF_SUCCESS}
;;
*)
;;
@@ -86,72 +111,111 @@ esac
ganesha_mon_start()
{
+ ocf_log debug "ganesha_mon_start"
+ ganesha_mon_monitor
return $OCF_SUCCESS
}
ganesha_mon_stop()
{
+ ocf_log debug "ganesha_mon_stop"
return $OCF_SUCCESS
}
ganesha_mon_monitor()
{
- local short_host=$(hostname -s)
- local pid_file="/var/run/ganesha.nfsd.pid"
+ local host=$(hostname -s)
+ local pid_file="/var/run/ganesha.pid"
+ local rhel6_pid_file="/var/run/ganesha.nfsd.pid"
+ local proc_pid="/proc/"
# RHEL6 /etc/init.d/nfs-ganesha adds -p /var/run/ganesha.nfsd.pid
# RHEL7 systemd does not. Would be nice if all distros used the
# same pid file.
- if [ -e /usr/lib/systemd/system/nfs-ganesha.service ]; then
- pid_file="/var/run/ganesha.pid"
+ if [ -e ${rhel6_pid_file} ]; then
+ pid_file=${rhel6_pid_file}
+ fi
+ if [ -e ${pid_file} ]; then
+ proc_pid="${proc_pid}$(cat ${pid_file})"
fi
- if [ -e ${pid_file} -a \
- -d /proc/$(cat ${pid_file} ) ]; then
- ( pcs resource delete ${short_host}-dead_ip-1 > /dev/null 2>&1 )
+ if [ "x${proc_pid}" != "x/proc/" -a -d ${proc_pid} ]; then
- attrd_updater -n ganesha-active -v 1
+ attrd_updater -n ${OCF_RESKEY_ganesha_active} -v 1
if [ $? -ne 0 ]; then
- logger "warning: attrd_updater -n ganesha-active -v 1 failed"
+ ocf_log info "warning: attrd_updater -n ${OCF_RESKEY_ganesha_active} -v 1 failed"
fi
- else
- ( pcs resource create ${short_host}-dead_ip-1 ocf:heartbeat:Dummy > /dev/null 2>&1 )
+ # 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
- logger "warning: pcs resource create ${short_host}-dead_ip-1 ocf:heartbeat:Dummy failed"
+ ocf_log info "warning: attrd_updater -n ${OCF_RESKEY_grace_active} -v 1 failed"
fi
- # The ${this-node}-dead_ip-1 resource is used to indicate
- # that this ganesha.nfsd has died.
- # VIP fail-over is then triggered by clearing the
- # ganesha-active node attribute on this node.
- #
- # Meanwhile the ganesha_grace monitor() runs every 5
- # seconds. We need to allow time for it to run and put
- # the remaining ganesha.nfsds into grace before initiating
- # the VIP fail-over.
- sleep ${GRACE_DELAY}
-
- attrd_updater -D -n ganesha-active
+ # ganesha_mon (nfs-mon) and ganesha_grace (nfs-grace)
+ # track grace-active crm_attr (attr != crm_attr)
+ # we can't just use the attr as there's no way to query
+ # its value in RHEL6 pacemaker
+
+ crm_attribute --node=${host} --lifetime=forever --name=${OCF_RESKEY_grace_active} --update=1 2> /dev/null
if [ $? -ne 0 ]; then
- logger "warning: attrd_updater -D -n ganesha-active failed"
+ host=$(hostname)
+ crm_attribute --node=${host} --lifetime=forever --name=${OCF_RESKEY_grace_active} --update=1 2> /dev/null
+ if [ $? -ne 0 ]; then
+ ocf_log info "mon monitor warning: crm_attribute --node=${host} --lifetime=forever --name=${OCF_RESKEY_grace_active} --update=1 failed"
+ fi
fi
+
+ return ${OCF_SUCCESS}
fi
- return $OCF_SUCCESS
+ # VIP fail-over is triggered by clearing the
+ # ganesha-active node attribute on this node.
+ #
+ # Meanwhile the ganesha_grace notify() runs when its
+ # nfs-grace resource is disabled on a node; which
+ # 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.
+
+ attrd_updater -D -n ${OCF_RESKEY_grace_active}
+ if [ $? -ne 0 ]; then
+ ocf_log info "warning: attrd_updater -D -n ${OCF_RESKEY_grace_active} failed"
+ fi
+
+ host=$(hostname -s)
+ crm_attribute --node=${host} --name=${OCF_RESKEY_grace_active} --update=0 2> /dev/null
+ if [ $? -ne 0 ]; then
+ host=$(hostname)
+ crm_attribute --node=${host} --name=${OCF_RESKEY_grace_active} --update=0 2> /dev/null
+ if [ $? -ne 0 ]; then
+ ocf_log info "mon monitor warning: crm_attribute --node=${host} --name=${OCF_RESKEY_grace_active} --update=0 failed"
+ fi
+ fi
+
+ sleep ${OCF_RESKEY_grace_delay}
+
+ attrd_updater -D -n ${OCF_RESKEY_ganesha_active}
+ if [ $? -ne 0 ]; then
+ ocf_log info "warning: attrd_updater -D -n ${OCF_RESKEY_ganesha_active} failed"
+ fi
+
+ return ${OCF_SUCCESS}
}
ganesha_mon_validate()
{
- return $OCF_SUCCESS
+ return ${OCF_SUCCESS}
}
ganesha_mon_validate
-# logger "ganesha_mon ${OCF_RESOURCE_INSTANCE} $__OCF_ACTION"
-
# Translate each action into the appropriate function call
-case $__OCF_ACTION in
+case ${__OCF_ACTION} in
start) ganesha_mon_start
;;
stop) ganesha_mon_stop
@@ -159,13 +223,12 @@ stop) ganesha_mon_stop
status|monitor) ganesha_mon_monitor
;;
*) ganesha_mon_usage
- exit $OCF_ERR_UNIMPLEMENTED
- ;;
+ exit ${OCF_ERR_UNIMPLEMENTED}
+ ;;
esac
rc=$?
# The resource agent may optionally log a debug message
-ocf_log debug "${OCF_RESOURCE_INSTANCE} $__OCF_ACTION returned $rc"
+ocf_log debug "${OCF_RESOURCE_INSTANCE} ${__OCF_ACTION} returned $rc"
exit $rc
-
diff --git a/extras/ganesha/ocf/ganesha_nfsd b/extras/ganesha/ocf/ganesha_nfsd
index e064183daef..f91e8b6b8f7 100644
--- a/extras/ganesha/ocf/ganesha_nfsd
+++ b/extras/ganesha/ocf/ganesha_nfsd
@@ -29,15 +29,18 @@
: ${OCF_FUNCTIONS_DIR=${OCF_ROOT}/lib/heartbeat}
. ${OCF_FUNCTIONS_DIR}/ocf-shellfuncs
-if [ -n "$OCF_DEBUG_LIBRARY" ]; then
- . $OCF_DEBUG_LIBRARY
+if [ -n "${OCF_DEBUG_LIBRARY}" ]; then
+ . ${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="/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">
@@ -59,16 +62,16 @@ resource agent for nfs-ganesha.
</parameters>
<actions>
-<action name="start" timeout="40s" />
-<action name="stop" timeout="40s" />
-<action name="status" depth="0" timeout="20s" interval="1m" />
-<action name="monitor" depth="0" timeout="10s" interval="1m" />
+<action name="start" timeout="5s" />
+<action name="stop" timeout="5s" />
+<action name="status" depth="0" timeout="5s" interval="0" />
+<action name="monitor" depth="0" timeout="5s" interval="0" />
<action name="meta-data" timeout="20s" />
</actions>
</resource-agent>
END
-return $OCF_SUCCESS
+return ${OCF_SUCCESS}
}
ganesha_nfsd_usage() {
@@ -78,10 +81,10 @@ ganesha_nfsd_usage() {
# Make sure meta-data and usage always succeed
case $__OCF_ACTION in
meta-data) ganesha_meta_data
- exit $OCF_SUCCESS
+ exit ${OCF_SUCCESS}
;;
usage|help) ganesha_usage
- exit $OCF_SUCCESS
+ exit ${OCF_SUCCESS}
;;
*)
;;
@@ -89,58 +92,60 @@ esac
ganesha_nfsd_start()
{
- return $OCF_SUCCESS
+ local long_host=$(hostname)
+
+ if [[ -d /var/lib/nfs ]]; then
+ mv /var/lib/nfs /var/lib/nfs.backup
+ if [ $? -ne 0 ]; then
+ ocf_log notice "mv /var/lib/nfs /var/lib/nfs.backup failed"
+ fi
+ ln -s ${OCF_RESKEY_ha_vol_mnt}/nfs-ganesha/${long_host}/nfs /var/lib/nfs
+ if [ $? -ne 0 ]; then
+ ocf_log notice "ln -s ${OCF_RESKEY_ha_vol_mnt}/nfs-ganesha/${long_host}/nfs /var/lib/nfs failed"
+ fi
+ fi
+
+ return ${OCF_SUCCESS}
}
ganesha_nfsd_stop()
{
- local short_host=$(hostname -s)
- local long_host=""
-
- if [ "X${OCF_RESOURCE_INSTANCE:0:9}X" = "Xnfs_startX" ]; then
-
- # if this is any nfs_start, go ahead. worst case we
- # find the link already exists and do nothing
- long_host=$(hostname)
-
- if [ -d /var/lib/nfs ]; then
- mv /var/lib/nfs /var/lib/nfs.backup
- ln -s $OCF_RESKEY_ha_vol_mnt/nfs-ganesha/${long_host}/nfs /var/lib/nfs
- if [ $? -ne 0 ]; then
- logger "warning: ln -s $OCF_RESKEY_ha_vol_mnt/nfs-ganesha/${long_host}/nfs /var/lib/nfs failed"
- fi
+ if [ -L /var/lib/nfs -a -d /var/lib/nfs.backup ]; then
+ rm -f /var/lib/nfs
+ if [ $? -ne 0 ]; then
+ ocf_log notice "rm -f /var/lib/nfs failed"
fi
- else
-
- # if this is a clone resource or is specific to this node
- # remove the symlink and restore /var/lib/nfs
-
- if [ "X${OCF_RESOURCE_INSTANCE}X" = "Xnfs_stopX" ] ||
- [ "X${OCF_RESOURCE_INSTANCE}X" = "Xnfs_stop-${short_host}X" ]; then
- if [ -L /var/lib/nfs -a -d /var/lib/nfs.backup ]; then
- rm -f /var/lib/nfs
- mv /var/lib/nfs.backup /var/lib/nfs
- fi
+ mv /var/lib/nfs.backup /var/lib/nfs
+ if [ $? -ne 0 ]; then
+ ocf_log notice "mv /var/lib/nfs.backup /var/lib/nfs failed"
fi
fi
- return $OCF_SUCCESS
+ return ${OCF_SUCCESS}
}
ganesha_nfsd_monitor()
{
- return $OCF_SUCCESS
+ # pacemaker checks to see if RA is already running before starting it.
+ # if we return success, then it's presumed it's already running and
+ # doesn't need to be started, i.e. invoke the start action.
+ # return something other than success to make pacemaker invoke the
+ # start action
+ if [[ -L /var/lib/nfs ]]; then
+ return ${OCF_SUCCESS}
+ fi
+ return ${OCF_NOT_RUNNING}
}
ganesha_nfsd_validate()
{
- return $OCF_SUCCESS
+ return ${OCF_SUCCESS}
}
ganesha_nfsd_validate
-# logger "ganesha_nfsd ${OCF_RESOURCE_INSTANCE} $__OCF_ACTION"
+# ocf_log notice "ganesha_nfsd ${OCF_RESOURCE_INSTANCE} $__OCF_ACTION"
# Translate each action into the appropriate function call
case $__OCF_ACTION in
@@ -151,13 +156,12 @@ stop) ganesha_nfsd_stop
status|monitor) ganesha_nfsd_monitor
;;
*) ganesha_nfsd_usage
- exit $OCF_ERR_UNIMPLEMENTED
- ;;
+ exit ${OCF_ERR_UNIMPLEMENTED}
+ ;;
esac
rc=$?
# The resource agent may optionally log a debug message
-ocf_log debug "${OCF_RESOURCE_INSTANCE} $__OCF_ACTION returned $rc"
+ocf_log debug "${OCF_RESOURCE_INSTANCE} ${__OCF_ACTION} returned $rc"
exit $rc
-
diff --git a/extras/ganesha/scripts/Makefile.am b/extras/ganesha/scripts/Makefile.am
index e71e2f6456b..7e345fd5f19 100644
--- a/extras/ganesha/scripts/Makefile.am
+++ b/extras/ganesha/scripts/Makefile.am
@@ -1,4 +1,6 @@
-EXTRA_DIST= ganesha-ha.sh dbus-send.sh create-export-ganesha.sh
+EXTRA_DIST= create-export-ganesha.sh generate-epoch.py dbus-send.sh \
+ ganesha-ha.sh
scriptsdir = $(libexecdir)/ganesha
-scripts_SCRIPTS = create-export-ganesha.sh dbus-send.sh ganesha-ha.sh
+scripts_SCRIPTS = create-export-ganesha.sh dbus-send.sh generate-epoch.py \
+ ganesha-ha.sh
diff --git a/extras/ganesha/scripts/create-export-ganesha.sh b/extras/ganesha/scripts/create-export-ganesha.sh
index ab7c282af79..3040e8138b0 100755
--- a/extras/ganesha/scripts/create-export-ganesha.sh
+++ b/extras/ganesha/scripts/create-export-ganesha.sh
@@ -1,20 +1,37 @@
-#/bin/bash
+#!/bin/bash
#This script is called by glusterd when the user
#tries to export a volume via NFS-Ganesha.
#An export file specific to a volume
#is created in GANESHA_DIR/exports.
+# Try loading the config from any of the distro
+# specific configuration locations
+if [ -f /etc/sysconfig/ganesha ]
+ then
+ . /etc/sysconfig/ganesha
+fi
+if [ -f /etc/conf.d/ganesha ]
+ then
+ . /etc/conf.d/ganesha
+fi
+if [ -f /etc/default/ganesha ]
+ then
+ . /etc/default/ganesha
+fi
+
GANESHA_DIR=${1%/}
-VOL=$2
-CONF=
-CONFFILE=
+OPTION=$2
+VOL=$3
+CONF=$GANESHA_DIR"/ganesha.conf"
+declare -i EXPORT_ID
function check_cmd_status()
{
if [ "$1" != "0" ]
then
rm -rf $GANESHA_DIR/exports/export.$VOL.conf
+ sed -i /$VOL.conf/d $CONF
exit 1
fi
}
@@ -26,34 +43,6 @@ if [ ! -d "$GANESHA_DIR/exports" ];
check_cmd_status `echo $?`
fi
-function find_rhel7_conf
-{
- while [[ $# > 0 ]]
- do
- key="$1"
- case $key in
- -f)
- CONFFILE="$2"
- ;;
- *)
- ;;
- esac
- shift
- done
-}
-
-cfgline=$(grep ^CONFFILE= /etc/sysconfig/ganesha)
-eval $(echo ${cfgline} | grep -F ^CONFFILE=)
-
-if [ -z $CONFFILE ]
- then
- cfgline=$(grep ^OPTIONS= /etc/sysconfig/ganesha)
- eval $(echo ${cfgline} | grep -F ^OPTIONS=)
- find_rhel7_conf $cfgline
-
-fi
-CONF=${CONFFILE:-/etc/ganesha/ganesha.conf}
-
function write_conf()
{
echo -e "# WARNING : Using Gluster CLI will overwrite manual
@@ -75,11 +64,29 @@ echo " Pseudo=\"/$VOL\";"
echo ' Protocols = "3", "4" ;'
echo ' Transports = "UDP","TCP";'
echo ' SecType = "sys";'
+echo ' Security_Label = False;'
echo " }"
}
-
-write_conf $@ > $GANESHA_DIR/exports/export.$VOL.conf
-if ! (cat $CONF | grep $VOL.conf$ )
+if [ "$OPTION" = "on" ];
then
-echo "%include \"$GANESHA_DIR/exports/export.$VOL.conf\"" >> $CONF
+ if ! (cat $CONF | grep $VOL.conf\"$ )
+ then
+ write_conf $@ > $GANESHA_DIR/exports/export.$VOL.conf
+ echo "%include \"$GANESHA_DIR/exports/export.$VOL.conf\"" >> $CONF
+ count=`ls -l $GANESHA_DIR/exports/*.conf | wc -l`
+ if [ "$count" = "1" ] ; then
+ EXPORT_ID=2
+ else
+ EXPORT_ID=`cat $GANESHA_DIR/.export_added`
+ check_cmd_status `echo $?`
+ EXPORT_ID=EXPORT_ID+1
+ sed -i s/Export_Id.*/"Export_Id= $EXPORT_ID ;"/ \
+ $GANESHA_DIR/exports/export.$VOL.conf
+ check_cmd_status `echo $?`
+ fi
+ echo $EXPORT_ID > $GANESHA_DIR/.export_added
+ fi
+else
+ rm -rf $GANESHA_DIR/exports/export.$VOL.conf
+ sed -i /$VOL.conf/d $CONF
fi
diff --git a/extras/ganesha/scripts/dbus-send.sh b/extras/ganesha/scripts/dbus-send.sh
index 5f268eb37c2..9d613a0e7ad 100755
--- a/extras/ganesha/scripts/dbus-send.sh
+++ b/extras/ganesha/scripts/dbus-send.sh
@@ -1,102 +1,62 @@
-#/bin/bash
+#!/bin/bash
-declare -i EXPORT_ID
-GANESHA_DIR=${1%/}
-OPTION=$2
-VOL=$3
-CONF=
-CONFFILE=
-
-function find_rhel7_conf
-{
- while [[ $# > 0 ]]
- do
- key="$1"
- case $key in
- -f)
- CONFFILE="$2"
- break;
- ;;
- *)
- ;;
- esac
- shift
- done
-}
-
-cfgline=$(grep ^CONFFILE= /etc/sysconfig/ganesha)
-eval $(echo ${cfgline} | grep -F ^CONFFILE=)
-
-if [ -z $CONFFILE ]
+# Try loading the config from any of the distro
+# specific configuration locations
+if [ -f /etc/sysconfig/ganesha ]
then
- cfgline=$(grep ^OPTIONS= /etc/sysconfig/ganesha)
- eval $(echo ${cfgline} | grep -F ^OPTIONS=)
- find_rhel7_conf $cfgline
-
+ . /etc/sysconfig/ganesha
+fi
+if [ -f /etc/conf.d/ganesha ]
+ then
+ . /etc/conf.d/ganesha
+fi
+if [ -f /etc/default/ganesha ]
+ then
+ . /etc/default/ganesha
fi
-CONF=${CONFFILE:-/etc/ganesha/ganesha.conf}
+GANESHA_DIR=${1%/}
+OPTION=$2
+VOL=$3
+CONF=$GANESHA_DIR"/ganesha.conf"
function check_cmd_status()
{
if [ "$1" != "0" ]
- then
- rm -rf $GANESHA_DIR/exports/export.$VOL.conf
- sed -i /$VOL.conf/d $CONF
- exit 1
+ then
+ logger "dynamic export failed on node :${hostname -s}"
fi
}
#This function keeps track of export IDs and increments it with every new entry
function dynamic_export_add()
{
- count=`ls -l $GANESHA_DIR/exports/*.conf | wc -l`
- if [ "$count" = "1" ] ;
- then
- EXPORT_ID=2
- else
- #if [ -s /var/lib/ganesha/export_removed ];
- # then
- # EXPORT_ID=`head -1 /var/lib/ganesha/export_removed`
- # sed -i -e "1d" /var/lib/ganesha/export_removed
- # else
-
- EXPORT_ID=`cat $GANESHA_DIR/.export_added`
- check_cmd_status `echo $?`
- #fi
- fi
- for entry in `grep -n Export_Id $GANESHA_DIR/exports/export.$VOL.conf \
- | awk -F":" '{print$1}'`
- do
- sed -e "$entry s/Export_Id.*/Export_Id=$EXPORT_ID ;/" -i \
- $GANESHA_DIR/exports/export.$VOL.conf
- check_cmd_status `echo $?`
- dbus-send --system \
- --dest=org.ganesha.nfsd /org/ganesha/nfsd/ExportMgr \
- org.ganesha.nfsd.exportmgr.AddExport \
- string:$GANESHA_DIR/exports/export.$VOL.conf \
- string:"EXPORT(Export_Id=$EXPORT_ID)"
- EXPORT_ID=EXPORT_ID+1
- done
- echo $EXPORT_ID > $GANESHA_DIR/.export_added
+ dbus-send --system \
+--dest=org.ganesha.nfsd /org/ganesha/nfsd/ExportMgr \
+org.ganesha.nfsd.exportmgr.AddExport string:$GANESHA_DIR/exports/export.$VOL.conf \
+string:"EXPORT(Path=/$VOL)"
check_cmd_status `echo $?`
}
#This function removes an export dynamically(uses the export_id of the export)
function dynamic_export_remove()
{
- grep Export_Id $GANESHA_DIR/exports/export.$VOL.conf | \
- while read entry;
- do
- dbus-send --print-reply --system \
- --dest=org.ganesha.nfsd /org/ganesha/nfsd/ExportMgr \
- org.ganesha.nfsd.exportmgr.RemoveExport \
- uint16:$(echo $entry | awk -F"[=,;]" '{print$2}')
- check_cmd_status `echo $?`
- done
- sed -i /$VOL.conf/d $CONF
- rm -rf $GANESHA_DIR/exports/export.$VOL.conf
+ # Below bash fetch all the export from ShowExport command and search
+ # export entry based on path and then get its export entry.
+ # There are two possiblities for path, either entire volume will be
+ # exported or subdir. It handles both cases. But it remove only first
+ # entry from the list based on assumption that entry exported via cli
+ # has lowest export id value
+ removed_id=$(dbus-send --type=method_call --print-reply --system \
+ --dest=org.ganesha.nfsd /org/ganesha/nfsd/ExportMgr \
+ org.ganesha.nfsd.exportmgr.ShowExports | grep -B 1 -we \
+ "/"$VOL -e "/"$VOL"/" | grep uint16 | awk '{print $2}' \
+ | head -1)
+ dbus-send --print-reply --system \
+--dest=org.ganesha.nfsd /org/ganesha/nfsd/ExportMgr \
+org.ganesha.nfsd.exportmgr.RemoveExport uint16:$removed_id
+ check_cmd_status `echo $?`
}
if [ "$OPTION" = "on" ];
@@ -108,4 +68,3 @@ if [ "$OPTION" = "off" ];
then
dynamic_export_remove $@
fi
-
diff --git a/extras/ganesha/scripts/ganesha-ha.sh b/extras/ganesha/scripts/ganesha-ha.sh
index f17582b727a..9790a719e10 100644
--- a/extras/ganesha/scripts/ganesha-ha.sh
+++ b/extras/ganesha/scripts/ganesha-ha.sh
@@ -1,6 +1,6 @@
#!/bin/bash
-# Copyright 2015 Red Hat Inc. All Rights Reserved
+# Copyright 2015-2016 Red Hat Inc. All Rights Reserved
#
# Pacemaker+Corosync High Availability for NFS-Ganesha
#
@@ -20,18 +20,42 @@
# ensure that the NFS GRACE DBUS signal is sent after the VIP moves to
# the new host.
+GANESHA_HA_SH=$(realpath $0)
HA_NUM_SERVERS=0
HA_SERVERS=""
-HA_CONFDIR="/etc/ganesha"
HA_VOL_NAME="gluster_shared_storage"
-HA_VOL_MNT="/var/run/gluster/shared_storage"
+HA_VOL_MNT="/run/gluster/shared_storage"
+HA_CONFDIR=$HA_VOL_MNT"/nfs-ganesha"
SERVICE_MAN="DISTRO_NOT_FOUND"
-RHEL6_PCS_CNAME_OPTION="--name"
+# rhel, fedora id, version
+ID=""
+VERSION_ID=""
+
+PCS9OR10_PCS_CNAME_OPTION=""
+PCS9OR10_PCS_CLONE_OPTION="clone"
SECRET_PEM="/var/lib/glusterd/nfs/secret.pem"
+# UNBLOCK RA uses shared_storage which may become unavailable
+# during any of the nodes reboot. Hence increase timeout value.
+PORTBLOCK_UNBLOCK_TIMEOUT="60s"
+
+# Try loading the config from any of the distro
+# specific configuration locations
+if [ -f /etc/sysconfig/ganesha ]
+ then
+ . /etc/sysconfig/ganesha
+fi
+if [ -f /etc/conf.d/ganesha ]
+ then
+ . /etc/conf.d/ganesha
+fi
+if [ -f /etc/default/ganesha ]
+ then
+ . /etc/default/ganesha
+fi
+
GANESHA_CONF=
-CONFFILE=
function find_rhel7_conf
{
@@ -50,14 +74,9 @@ function find_rhel7_conf
done
}
-cfgline=$(grep ^CONFFILE= /etc/sysconfig/ganesha)
-eval $(echo ${cfgline} | grep -F ^CONFFILE=)
-
-if [ -z $CONFFILE ]
+if [ -z ${CONFFILE} ]
then
- cfgline=$(grep ^OPTIONS= /etc/sysconfig/ganesha)
- eval $(echo ${cfgline} | grep -F ^OPTIONS=)
- find_rhel7_conf $cfgline
+ find_rhel7_conf ${OPTIONS}
fi
@@ -65,20 +84,21 @@ GANESHA_CONF=${CONFFILE:-/etc/ganesha/ganesha.conf}
usage() {
- echo "Usage : add|delete|status"
- echo "Add-node : ganesha-ha.sh --add <HA_CONF_DIR> \
+ echo "Usage : add|delete|refresh-config|status"
+ echo "Add-node : ganesha-ha.sh --add <HA_CONF_DIR> \
<NODE-HOSTNAME> <NODE-VIP>"
- echo "Delete-node: ganesha-ha.sh --delete <HA_CONF_DIR> \
+ echo "Delete-node: ganesha-ha.sh --delete <HA_CONF_DIR> \
<NODE-HOSTNAME>"
- echo "Refresh-config : ganesha-ha.sh --refresh-config <HA_CONFDIR>\
- <volume>"
+ echo "Refresh-config : ganesha-ha.sh --refresh-config <HA_CONFDIR> \
+<volume>"
+ echo "Status : ganesha-ha.sh --status <HA_CONFDIR>"
}
determine_service_manager () {
- if [ -e "/usr/bin/systemctl" ];
+ if [ -e "/bin/systemctl" ];
then
- SERVICE_MAN="/usr/bin/systemctl"
+ SERVICE_MAN="/bin/systemctl"
elif [ -e "/sbin/invoke-rc.d" ];
then
SERVICE_MAN="/sbin/invoke-rc.d"
@@ -86,9 +106,9 @@ determine_service_manager () {
then
SERVICE_MAN="/sbin/service"
fi
- if [ "$SERVICE_MAN" == "DISTRO_NOT_FOUND" ]
+ if [[ "${SERVICE_MAN}X" == "DISTRO_NOT_FOUNDX" ]]
then
- echo "Service manager not recognized, exiting"
+ logger "Service manager not recognized, exiting"
exit 1
fi
}
@@ -97,16 +117,27 @@ manage_service ()
{
local action=${1}
local new_node=${2}
- if [ "$SERVICE_MAN" == "/usr/sbin/systemctl" ]
+ local option=
+
+ if [[ "${action}" == "start" ]]; then
+ option="yes"
+ else
+ option="no"
+ fi
+ ssh -oPasswordAuthentication=no -oStrictHostKeyChecking=no -i \
+${SECRET_PEM} root@${new_node} "${GANESHA_HA_SH} --setup-ganesha-conf-files $HA_CONFDIR $option"
+
+ if [[ "${SERVICE_MAN}" == "/bin/systemctl" ]]
then
ssh -oPasswordAuthentication=no -oStrictHostKeyChecking=no -i \
-${SECRET_PEM} root@${new_node} "$SERVICE_MAN ${action} nfs-ganesha"
+${SECRET_PEM} root@${new_node} "${SERVICE_MAN} ${action} nfs-ganesha"
else
ssh -oPasswordAuthentication=no -oStrictHostKeyChecking=no -i \
-${SECRET_PEM} root@${new_node} "$SERVICE_MAN nfs-ganesha ${action}"
+${SECRET_PEM} root@${new_node} "${SERVICE_MAN} nfs-ganesha ${action}"
fi
}
+
check_cluster_exists()
{
local name=${1}
@@ -114,7 +145,7 @@ check_cluster_exists()
if [ -e /var/run/corosync.pid ]; then
cluster_name=$(pcs status | grep "Cluster name:" | cut -d ' ' -f 3)
- if [ ${cluster_name} -a ${cluster_name} = ${name} ]; then
+ if [[ "${cluster_name}X" == "${name}X" ]]; then
logger "$name already exists, exiting"
exit 0
fi
@@ -129,7 +160,7 @@ determine_servers()
local tmp_ifs=${IFS}
local ha_servers=""
- if [[ "X${cmd}X" != "XsetupX" ]]; then
+ if [ "${cmd}X" != "setupX" -a "${cmd}X" != "statusX" ]; then
ha_servers=$(pcs status | grep "Online:" | grep -o '\[.*\]' | sed -e 's/\[//' | sed -e 's/\]//')
IFS=$' '
for server in ${ha_servers} ; do
@@ -149,6 +180,13 @@ determine_servers()
fi
}
+stop_ganesha_all()
+{
+ local serverlist=${1}
+ for node in ${serverlist} ; do
+ manage_service "stop" ${node}
+ done
+}
setup_cluster()
{
@@ -156,37 +194,54 @@ setup_cluster()
local num_servers=${2}
local servers=${3}
local unclean=""
+ local quorum_policy="stop"
logger "setting up cluster ${name} with the following ${servers}"
- pcs cluster auth ${servers}
-# fedora pcs cluster setup ${name} ${servers}
-# rhel6 pcs cluster setup --name ${name} ${servers}
- pcs cluster setup ${RHEL6_PCS_CNAME_OPTION} ${name} ${servers}
+ # pcs cluster setup --force ${PCS9OR10_PCS_CNAME_OPTION} ${name} ${servers}
+ pcs cluster setup --force ${PCS9OR10_PCS_CNAME_OPTION} ${name} --enable ${servers}
if [ $? -ne 0 ]; then
- logger "pcs cluster setup ${RHEL6_PCS_CNAME_OPTION} ${name} ${servers} failed"
+ logger "pcs cluster setup ${PCS9OR10_PCS_CNAME_OPTION} ${name} --enable ${servers} failed, shutting down ganesha and bailing out"
+ #set up failed stop all ganesha process and clean up symlinks in cluster
+ stop_ganesha_all "${servers}"
exit 1;
fi
+
+ # pcs cluster auth ${servers}
+ pcs cluster auth
+ if [ $? -ne 0 ]; then
+ logger "pcs cluster auth failed"
+ fi
+
pcs cluster start --all
if [ $? -ne 0 ]; then
logger "pcs cluster start failed"
exit 1;
fi
- sleep 3
+ sleep 1
+ # wait for the cluster to elect a DC before querying or writing
+ # to the CIB. BZ 1334092
+ crmadmin --dc_lookup --timeout=5000 > /dev/null 2>&1
+ while [ $? -ne 0 ]; do
+ crmadmin --dc_lookup --timeout=5000 > /dev/null 2>&1
+ done
+
unclean=$(pcs status | grep -u "UNCLEAN")
- while [[ "${unclean}X" = "UNCLEANX" ]]; do
+ while [[ "${unclean}X" == "UNCLEANX" ]]; do
sleep 1
unclean=$(pcs status | grep -u "UNCLEAN")
done
sleep 1
if [ ${num_servers} -lt 3 ]; then
- pcs property set no-quorum-policy=ignore
- if [ $? -ne 0 ]; then
- logger "warning: pcs property set no-quorum-policy=ignore failed"
- fi
+ quorum_policy="ignore"
+ fi
+ pcs property set no-quorum-policy=${quorum_policy}
+ if [ $? -ne 0 ]; then
+ logger "warning: pcs property set no-quorum-policy=${quorum_policy} failed"
fi
+
pcs property set stonith-enabled=false
if [ $? -ne 0 ]; then
logger "warning: pcs property set stonith-enabled=false failed"
@@ -194,123 +249,68 @@ setup_cluster()
}
-setup_finalize()
+setup_finalize_ha()
{
local cibfile=${1}
local stopped=""
stopped=$(pcs status | grep -u "Stopped")
- while [[ "${stopped}X" = "StoppedX" ]]; do
+ while [[ "${stopped}X" == "StoppedX" ]]; do
sleep 1
stopped=$(pcs status | grep -u "Stopped")
done
-
- pcs status | grep dead_ip-1 | sort > /var/run/ganesha/pcs_status
-
}
-setup_copy_config()
-{
- local short_host=$(hostname -s)
- local tganesha_conf=$(mktemp -u)
-
- if [ -e ${SECRET_PEM} ]; then
- while [[ ${1} ]]; do
- current_host=`echo ${1} | cut -d "." -f 1`
- if [ ${short_host} != ${current_host} ]; then
- scp -oPasswordAuthentication=no -oStrictHostKeyChecking=no -i \
-${SECRET_PEM} ${HA_CONFDIR}/ganesha-ha.conf ${1}:${HA_CONFDIR}/
- if [ $? -ne 0 ]; then
- logger "warning: scp ganesha-ha.conf to ${1} failed"
- fi
- fi
- shift
- done
- else
- logger "warning: scp ganesha-ha.conf to ${1} failed"
- fi
-}
-
refresh_config ()
{
local short_host=$(hostname -s)
local VOL=${1}
local HA_CONFDIR=${2}
+ local short_host=$(hostname -s)
+
+ local export_id=$(grep ^[[:space:]]*Export_Id $HA_CONFDIR/exports/export.$VOL.conf |\
+ awk -F"[=,;]" '{print $2}' | tr -d '[[:space:]]')
- removed_id=`cat $HA_CONFDIR/exports/export.$VOL.conf |\
-grep Export_Id | cut -d " " -f8`
if [ -e ${SECRET_PEM} ]; then
while [[ ${3} ]]; do
current_host=`echo ${3} | cut -d "." -f 1`
- if [ ${short_host} != ${current_host} ]; then
- scp -oPasswordAuthentication=no -oStrictHostKeyChecking=no -i \
-${SECRET_PEM} ${HA_CONFDIR}/exports/export.$VOL.conf \
-${current_host}:${HA_CONFDIR}/exports/
- grep Export_Id $HA_CONFDIR/exports/export.$VOL.conf | \
- while read entry;
- do
- export_id=$(echo $entry | awk -F"[=,;]" '{print$2}')
- ssh -oPasswordAuthentication=no \
--oStrictHostKeyChecking=no -i ${SECRET_PEM} root@${current_host} \
-"dbus-send --print-reply --system --dest=org.ganesha.nfsd \
-/org/ganesha/nfsd/ExportMgr org.ganesha.nfsd.exportmgr.RemoveExport \
-uint16:$export_id"
- sleep 1
- ssh -oPasswordAuthentication=no \
+ if [[ ${short_host} != ${current_host} ]]; then
+ output=$(ssh -oPasswordAuthentication=no \
-oStrictHostKeyChecking=no -i ${SECRET_PEM} root@${current_host} \
-"dbus-send --system --dest=org.ganesha.nfsd /org/ganesha/nfsd/ExportMgr \
-org.ganesha.nfsd.exportmgr.AddExport string:$HA_CONFDIR/exports/export.$VOL.conf \
-string:\"EXPORT(Export_Id=$export_id)\""
- if [ $? -ne 0 ]; then
- echo "warning: refresh-config failed on" \
- " ${current_host}"
- fi
- done
- fi
- shift
+"dbus-send --print-reply --system --dest=org.ganesha.nfsd \
+/org/ganesha/nfsd/ExportMgr org.ganesha.nfsd.exportmgr.UpdateExport \
+string:$HA_CONFDIR/exports/export.$VOL.conf \
+string:\"EXPORT(Export_Id=$export_id)\" 2>&1")
+ ret=$?
+ logger <<< "${output}"
+ if [ ${ret} -ne 0 ]; then
+ echo "Refresh-config failed on ${current_host}. Please check logs on ${current_host}"
+ else
+ echo "Refresh-config completed on ${current_host}."
+ fi
+
+ fi
+ shift
done
else
- echo "warning: refresh-config failed on ${1}"
+ echo "Error: refresh-config failed. Passwordless ssh is not enabled."
+ exit 1
fi
-#Run the same command on the localhost,
- grep Export_Id $HA_CONFDIR/exports/export.$VOL.conf | \
- while read entry;
- do
- export_id=$(echo $entry | awk -F"[=,;]" '{print$2}')
- dbus-send --print-reply --system \
- --dest=org.ganesha.nfsd /org/ganesha/nfsd/ExportMgr \
- org.ganesha.nfsd.exportmgr.RemoveExport uint16:$removed_id
- sleep 1
- dbus-send --system \
- --dest=org.ganesha.nfsd /org/ganesha/nfsd/ExportMgr \
- org.ganesha.nfsd.exportmgr.AddExport \
- string:$HA_CONFDIR/exports/export.$VOL.conf \
- string:"EXPORT(Export_Id=$export_id)"
- done
-}
-
-copy_export_config ()
-{
- local new_node=${1}
- local tganesha_conf=$(mktemp)
- local tganesha_exports=$(mktemp -d)
- local short_host=$(hostname -s)
- # avoid prompting for password, even with password-less scp
- # scp $host1:$file $host2:$file prompts for the password
- scp -oPasswordAuthentication=no -oStrictHostKeyChecking=no -i \
-${SECRET_PEM} ${HA_VOL_SERVER}:${GANESHA_CONF} $short_host:${tganesha_conf}
- scp -oPasswordAuthentication=no -oStrictHostKeyChecking=no -i \
-${SECRET_PEM} ${tganesha_conf} ${new_node}:${GANESHA_CONF}
- rm -f ${tganesha_conf}
-
- scp -r -oPasswordAuthentication=no -oStrictHostKeyChecking=no -i \
-${SECRET_PEM} ${HA_VOL_SERVER}:${HA_CONFDIR}/exports/ $short_host:${tganesha_exports}
- scp -r -oPasswordAuthentication=no -oStrictHostKeyChecking=no -i \
-${SECRET_PEM} ${tganesha_exports}/exports ${new_node}:${HA_CONFDIR}/
- rm -rf ${tganesha_exports}
+ # Run the same command on the localhost,
+ output=$(dbus-send --print-reply --system --dest=org.ganesha.nfsd \
+/org/ganesha/nfsd/ExportMgr org.ganesha.nfsd.exportmgr.UpdateExport \
+string:$HA_CONFDIR/exports/export.$VOL.conf \
+string:"EXPORT(Export_Id=$export_id)" 2>&1)
+ ret=$?
+ logger <<< "${output}"
+ if [ ${ret} -ne 0 ] ; then
+ echo "Refresh-config failed on localhost."
+ else
+ echo "Success: refresh-config completed."
+ fi
}
@@ -322,7 +322,7 @@ 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
@@ -334,13 +334,13 @@ teardown_cluster()
fi
done
-# BZ 1193433 - pcs doesn't reload cluster.conf after modification
-# after teardown completes, a subsequent setup will appear to have
-# 'remembered' the deleted node. You can work around this by
-# issuing another `pcs cluster node remove $node`,
-# `crm_node -f -R $server`, or
-# `cibadmin --delete --xml-text '<node id="$server"
-# uname="$server"/>'
+ # BZ 1193433 - pcs doesn't reload cluster.conf after modification
+ # after teardown completes, a subsequent setup will appear to have
+ # 'remembered' the deleted node. You can work around this by
+ # issuing another `pcs cluster node remove $node`,
+ # `crm_node -f -R $server`, or
+ # `cibadmin --delete --xml-text '<node id="$server"
+ # uname="$server"/>'
pcs cluster stop --all
if [ $? -ne 0 ]; then
@@ -357,11 +357,10 @@ teardown_cluster()
cleanup_ganesha_config ()
{
- rm -rf ${HA_CONFDIR}/exports/*.conf
- rm -rf ${HA_CONFDIR}/.export_added
- rm -rf /etc/cluster/cluster.conf*
- rm -rf /var/lib/pacemaker/cib/*
- sed -r -i -e '/^%include[[:space:]]+".+\.conf"$/d' ${GANESHA_CONF}
+ rm -f /etc/corosync/corosync.conf
+ rm -rf /etc/cluster/cluster.conf*
+ rm -rf /var/lib/pacemaker/cib/*
+ sed -r -i -e '/^%include[[:space:]]+".+\.conf"$/d' $HA_CONFDIR/ganesha.conf
}
do_create_virt_ip_constraints()
@@ -372,17 +371,17 @@ do_create_virt_ip_constraints()
# first a constraint location rule that says the VIP must be where
# there's a ganesha.nfsd running
- pcs -f ${cibfile} constraint location ${primary}-cluster_ip-1 rule score=-INFINITY ganesha-active ne 1
+ pcs -f ${cibfile} constraint location ${primary}-group rule score=-INFINITY ganesha-active ne 1
if [ $? -ne 0 ]; then
- logger "warning: pcs constraint location ${primary}-cluster_ip-1 rule score=-INFINITY ganesha-active ne 1 failed"
+ logger "warning: pcs constraint location ${primary}-group rule score=-INFINITY ganesha-active ne 1 failed"
fi
# then a set of constraint location prefers to set the prefered order
# for where a VIP should move
while [[ ${1} ]]; do
- pcs -f ${cibfile} constraint location ${primary}-cluster_ip-1 prefers ${1}=${weight}
+ pcs -f ${cibfile} constraint location ${primary}-group prefers ${1}=${weight}
if [ $? -ne 0 ]; then
- logger "warning: pcs constraint location ${primary}-cluster_ip-1 prefers ${1}=${weight} failed"
+ logger "warning: pcs constraint location ${primary}-group prefers ${1}=${weight} failed"
fi
weight=$(expr ${weight} + 1000)
shift
@@ -392,9 +391,9 @@ do_create_virt_ip_constraints()
# on Fedora setting appears to be additive, so to get the desired
# value we adjust the weight
# weight=$(expr ${weight} - 100)
- pcs -f ${cibfile} constraint location ${primary}-cluster_ip-1 prefers ${primary}=${weight}
+ pcs -f ${cibfile} constraint location ${primary}-group prefers ${primary}=${weight}
if [ $? -ne 0 ]; then
- logger "warning: pcs constraint location ${primary}-cluster_ip-1 prefers ${primary}=${weight} failed"
+ logger "warning: pcs constraint location ${primary}-group prefers ${primary}=${weight} failed"
fi
}
@@ -410,7 +409,7 @@ wrap_create_virt_ip_constraints()
# the result is "node2 node3 node4"; for node2, "node3 node4 node1"
# and so on.
while [[ ${1} ]]; do
- if [ "${1}" = "${primary}" ]; then
+ if [[ ${1} == ${primary} ]]; then
shift
while [[ ${1} ]]; do
tail=${tail}" "${1}
@@ -440,28 +439,31 @@ setup_create_resources()
{
local cibfile=$(mktemp -u)
- # mount the HA-state volume and start ganesha.nfsd on all nodes
- pcs resource create nfs_start ganesha_nfsd ha_vol_mnt=${HA_VOL_MNT} --clone
+ # fixup /var/lib/nfs
+ logger "pcs resource create nfs_setup ocf:heartbeat:ganesha_nfsd ha_vol_mnt=${HA_VOL_MNT} ${PCS9OR10_PCS_CLONE_OPTION}"
+ pcs resource create nfs_setup ocf:heartbeat:ganesha_nfsd ha_vol_mnt=${HA_VOL_MNT} ${PCS9OR10_PCS_CLONE_OPTION}
if [ $? -ne 0 ]; then
- logger "warning: pcs resource create nfs_start ganesha_nfsd ha_vol_mnt=${HA_VOL_MNT} --clone failed"
+ logger "warning: pcs resource create nfs_setup ocf:heartbeat:ganesha_nfsd ha_vol_mnt=${HA_VOL_MNT} ${PCS9OR10_PCS_CLONE_OPTION} failed"
fi
- sleep 1
- # cloned resources seem to never have their start() invoked when they
- # are created, but stop() is invoked when they are destroyed. Why???.
- # No matter, we don't want this resource agent hanging around anyway
- pcs resource delete nfs_start-clone
+
+ pcs resource create nfs-mon ocf:heartbeat:ganesha_mon ${PCS9OR10_PCS_CLONE_OPTION}
if [ $? -ne 0 ]; then
- logger "warning: pcs resource delete nfs_start-clone failed"
+ logger "warning: pcs resource create nfs-mon ocf:heartbeat:ganesha_mon ${PCS9OR10_PCS_CLONE_OPTION} failed"
fi
- pcs resource create nfs-mon ganesha_mon --clone
+ # see comment in (/usr/lib/ocf/resource.d/heartbeat/ganesha_grace
+ # start method. Allow time for ganesha_mon to start and set the
+ # ganesha-active crm_attribute
+ sleep 5
+
+ pcs resource create nfs-grace ocf:heartbeat:ganesha_grace ${PCS9OR10_PCS_CLONE_OPTION} notify=true
if [ $? -ne 0 ]; then
- logger "warning: pcs resource create nfs-mon ganesha_mon --clone failed"
+ logger "warning: pcs resource create nfs-grace ocf:heartbeat:ganesha_grace ${PCS9OR10_PCS_CLONE_OPTION} failed"
fi
- pcs resource create nfs-grace ganesha_grace --clone
+ pcs constraint location nfs-grace-clone rule score=-INFINITY grace-active ne 1
if [ $? -ne 0 ]; then
- logger "warning: pcs resource create nfs-grace ganesha_grace --clone failed"
+ logger "warning: pcs constraint location nfs-grace-clone rule score=-INFINITY grace-active ne 1"
fi
pcs cluster cib ${cibfile}
@@ -486,30 +488,32 @@ setup_create_resources()
eval tmp_ipaddr=\$${clean_name}
ipaddr=${tmp_ipaddr//_/.}
- pcs -f ${cibfile} resource create ${1}-cluster_ip-1 ocf:heartbeat:IPaddr ip=${ipaddr} cidr_netmask=32 op monitor interval=15s
+ pcs -f ${cibfile} resource create ${1}-nfs_block ocf:heartbeat:portblock protocol=tcp \
+ portno=2049 action=block ip=${ipaddr} --group ${1}-group
if [ $? -ne 0 ]; then
- logger "warning pcs resource create ${1}-cluster_ip-1 ocf:heartbeat:IPaddr ip=${ipaddr} cidr_netmask=32 op monitor interval=15s failed"
+ logger "warning pcs resource create ${1}-nfs_block failed"
fi
-
- pcs -f ${cibfile} resource create ${1}-trigger_ip-1 ocf:heartbeat:Dummy
+ pcs -f ${cibfile} resource create ${1}-cluster_ip-1 ocf:heartbeat:IPaddr ip=${ipaddr} \
+ cidr_netmask=32 op monitor interval=15s --group ${1}-group --after ${1}-nfs_block
if [ $? -ne 0 ]; then
- logger "warning: pcs resource create ${1}-trigger_ip-1 ocf:heartbeat:Dummy failed"
+ logger "warning pcs resource create ${1}-cluster_ip-1 ocf:heartbeat:IPaddr ip=${ipaddr} \
+ cidr_netmask=32 op monitor interval=15s failed"
fi
- pcs -f ${cibfile} constraint colocation add ${1}-cluster_ip-1 with ${1}-trigger_ip-1
+ pcs -f ${cibfile} constraint order nfs-grace-clone then ${1}-cluster_ip-1
if [ $? -ne 0 ]; then
- logger "warning: pcs constraint colocation add ${1}-cluster_ip-1 with ${1}-trigger_ip-1 failed"
+ logger "warning: pcs constraint order nfs-grace-clone then ${1}-cluster_ip-1 failed"
fi
- pcs -f ${cibfile} constraint order ${1}-trigger_ip-1 then nfs-grace-clone
+ pcs -f ${cibfile} resource create ${1}-nfs_unblock ocf:heartbeat:portblock protocol=tcp \
+ portno=2049 action=unblock ip=${ipaddr} reset_local_on_unblock_stop=true \
+ tickle_dir=${HA_VOL_MNT}/nfs-ganesha/tickle_dir/ --group ${1}-group --after ${1}-cluster_ip-1 \
+ op stop timeout=${PORTBLOCK_UNBLOCK_TIMEOUT} op start timeout=${PORTBLOCK_UNBLOCK_TIMEOUT} \
+ op monitor interval=10s timeout=${PORTBLOCK_UNBLOCK_TIMEOUT}
if [ $? -ne 0 ]; then
- logger "warning: pcs constraint order ${1}-trigger_ip-1 then nfs-grace-clone failed"
+ logger "warning pcs resource create ${1}-nfs_unblock failed"
fi
- pcs -f ${cibfile} constraint order nfs-grace-clone then ${1}-cluster_ip-1
- if [ $? -ne 0 ]; then
- logger "warning: pcs constraint order nfs-grace-clone then ${1}-cluster_ip-1 failed"
- fi
shift
done
@@ -528,6 +532,13 @@ teardown_resources()
{
# local mntpt=$(grep ha-vol-mnt ${HA_CONFIG_FILE} | cut -d = -f 2)
+ # restore /var/lib/nfs
+ logger "notice: pcs resource delete nfs_setup-clone"
+ pcs resource delete nfs_setup-clone
+ if [ $? -ne 0 ]; then
+ logger "warning: pcs resource delete nfs_setup-clone failed"
+ fi
+
# delete -clone resource agents
# in particular delete the ganesha monitor so we don't try to
# trigger anything when we shut down ganesha next.
@@ -541,31 +552,10 @@ teardown_resources()
logger "warning: pcs resource delete nfs-grace-clone failed"
fi
- # unmount the HA-state volume and terminate ganesha.nfsd on all nodes
- pcs resource create nfs_stop ganesha_nfsd ha_vol_mnt=${HA_VOL_MNT} --clone
- if [ $? -ne 0 ]; then
- logger "warning: pcs resource create nfs_stop ganesha_nfsd ha_vol_mnt=${HA_VOL_MNT} --clone failed"
- fi
- sleep 1
- # cloned resources seem to never have their start() invoked when they
- # are created, but stop() is invoked when they are destroyed. Why???.
- pcs resource delete nfs_stop-clone
- if [ $? -ne 0 ]; then
- logger "warning: pcs resource delete nfs_stop-clone failed"
- fi
-
while [[ ${1} ]]; do
- pcs resource delete ${1}-cluster_ip-1
- if [ $? -ne 0 ]; then
- logger "warning: pcs resource delete ${1}-cluster_ip-1 failed"
- fi
- pcs resource delete ${1}-trigger_ip-1
- if [ $? -ne 0 ]; then
- logger "warning: pcs resource delete ${1}-trigger_ip-1 failed"
- fi
- pcs resource delete ${1}-dead_ip-1
+ pcs resource delete ${1}-group
if [ $? -ne 0 ]; then
- logger "info: pcs resource delete ${1}-dead_ip-1 failed"
+ logger "warning: pcs resource delete ${1}-group failed"
fi
shift
done
@@ -588,29 +578,30 @@ recreate_resources()
eval tmp_ipaddr=\$${clean_name}
ipaddr=${tmp_ipaddr//_/.}
- pcs -f ${cibfile} resource create ${1}-cluster_ip-1 ocf:heartbeat:IPaddr ip=${ipaddr} cidr_netmask=32 op monitor interval=15s
- if [ $? -ne 0 ]; then
- logger "warning pcs resource create ${1}-cluster_ip-1 ocf:heartbeat:IPaddr ip=${ipaddr} cidr_netmask=32 op monitor interval=10s failed"
- fi
-
- pcs -f ${cibfile} resource create ${1}-trigger_ip-1 ocf:heartbeat:Dummy
+ pcs -f ${cibfile} resource create ${1}-nfs_block ocf:heartbeat:portblock protocol=tcp \
+ portno=2049 action=block ip=${ipaddr} --group ${1}-group
if [ $? -ne 0 ]; then
- logger "warning: pcs resource create ${1}-trigger_ip-1 ocf:heartbeat:Dummy failed"
+ logger "warning pcs resource create ${1}-nfs_block failed"
fi
-
- pcs -f ${cibfile} constraint colocation add ${1}-cluster_ip-1 with ${1}-trigger_ip-1
+ pcs -f ${cibfile} resource create ${1}-cluster_ip-1 ocf:heartbeat:IPaddr ip=${ipaddr} \
+ cidr_netmask=32 op monitor interval=15s --group ${1}-group --after ${1}-nfs_block
if [ $? -ne 0 ]; then
- logger "warning: pcs constraint colocation add ${1}-cluster_ip-1 with ${1}-trigger_ip-1 failed"
+ logger "warning pcs resource create ${1}-cluster_ip-1 ocf:heartbeat:IPaddr ip=${ipaddr} \
+ cidr_netmask=32 op monitor interval=15s failed"
fi
- pcs -f ${cibfile} constraint order ${1}-trigger_ip-1 then nfs-grace-clone
+ pcs -f ${cibfile} constraint order nfs-grace-clone then ${1}-cluster_ip-1
if [ $? -ne 0 ]; then
- logger "warning: pcs constraint order ${1}-trigger_ip-1 then nfs-grace-clone failed"
+ logger "warning: pcs constraint order nfs-grace-clone then ${1}-cluster_ip-1 failed"
fi
- pcs -f ${cibfile} constraint order nfs-grace-clone then ${1}-cluster_ip-1
+ pcs -f ${cibfile} resource create ${1}-nfs_unblock ocf:heartbeat:portblock protocol=tcp \
+ portno=2049 action=unblock ip=${ipaddr} reset_local_on_unblock_stop=true \
+ tickle_dir=${HA_VOL_MNT}/nfs-ganesha/tickle_dir/ --group ${1}-group --after ${1}-cluster_ip-1 \
+ op stop timeout=${PORTBLOCK_UNBLOCK_TIMEOUT} op start timeout=${PORTBLOCK_UNBLOCK_TIMEOUT} \
+ op monitor interval=10s timeout=${PORTBLOCK_UNBLOCK_TIMEOUT}
if [ $? -ne 0 ]; then
- logger "warning: pcs constraint order nfs-grace-clone then ${1}-cluster_ip-1 failed"
+ logger "warning pcs resource create ${1}-nfs_unblock failed"
fi
shift
@@ -626,30 +617,32 @@ addnode_recreate_resources()
recreate_resources ${cibfile} ${HA_SERVERS}
- pcs -f ${cibfile} resource create ${add_node}-cluster_ip-1 ocf:heartbeat:IPaddr ip=${add_vip} cidr_netmask=32 op monitor interval=15s
+ pcs -f ${cibfile} resource create ${add_node}-nfs_block ocf:heartbeat:portblock \
+ protocol=tcp portno=2049 action=block ip=${add_vip} --group ${add_node}-group
if [ $? -ne 0 ]; then
- logger "warning pcs resource create ${add_node}-cluster_ip-1 ocf:heartbeat:IPaddr ip=${add_vip} cidr_netmask=32 op monitor interval=10s failed"
+ logger "warning pcs resource create ${add_node}-nfs_block failed"
fi
-
- pcs -f ${cibfile} resource create ${add_node}-trigger_ip-1 ocf:heartbeat:Dummy
- if [ $? -ne 0 ]; then
- logger "warning: pcs resource create ${add_node}-trigger_ip-1 ocf:heartbeat:Dummy failed"
- fi
-
- pcs -f ${cibfile} constraint colocation add ${add_node}-cluster_ip-1 with ${add_node}-trigger_ip-1
+ pcs -f ${cibfile} resource create ${add_node}-cluster_ip-1 ocf:heartbeat:IPaddr \
+ ip=${add_vip} cidr_netmask=32 op monitor interval=15s --group ${add_node}-group \
+ --after ${add_node}-nfs_block
if [ $? -ne 0 ]; then
- logger "warning: pcs constraint colocation add ${add_node}-cluster_ip-1 with ${add_node}-trigger_ip-1 failed"
- fi
-
- pcs -f ${cibfile} constraint order ${add_node}-trigger_ip-1 then nfs-grace-clone
- if [ $? -ne 0 ]; then
- logger "warning: pcs constraint order ${add_node}-trigger_ip-1 then nfs-grace-clone failed"
+ logger "warning pcs resource create ${add_node}-cluster_ip-1 ocf:heartbeat:IPaddr \
+ ip=${add_vip} cidr_netmask=32 op monitor interval=15s failed"
fi
pcs -f ${cibfile} constraint order nfs-grace-clone then ${add_node}-cluster_ip-1
if [ $? -ne 0 ]; then
logger "warning: pcs constraint order nfs-grace-clone then ${add_node}-cluster_ip-1 failed"
fi
+ pcs -f ${cibfile} resource create ${add_node}-nfs_unblock ocf:heartbeat:portblock \
+ protocol=tcp portno=2049 action=unblock ip=${add_vip} reset_local_on_unblock_stop=true \
+ tickle_dir=${HA_VOL_MNT}/nfs-ganesha/tickle_dir/ --group ${add_node}-group --after \
+ ${add_node}-cluster_ip-1 op stop timeout=${PORTBLOCK_UNBLOCK_TIMEOUT} op start \
+ timeout=${PORTBLOCK_UNBLOCK_TIMEOUT} op monitor interval=10s \
+ timeout=${PORTBLOCK_UNBLOCK_TIMEOUT}
+ if [ $? -ne 0 ]; then
+ logger "warning pcs resource create ${add_node}-nfs_unblock failed"
+ fi
}
@@ -658,14 +651,9 @@ clear_resources()
local cibfile=${1}; shift
while [[ ${1} ]]; do
- pcs -f ${cibfile} resource delete ${1}-cluster_ip-1
- if [ $? -ne 0 ]; then
- logger "warning: pcs -f ${cibfile} resource delete ${1}-cluster_ip-1"
- fi
-
- pcs -f ${cibfile} resource delete ${1}-trigger_ip-1
+ pcs -f ${cibfile} resource delete ${1}-group
if [ $? -ne 0 ]; then
- logger "warning: pcs -f ${cibfile} resource delete ${1}-trigger_ip-1"
+ logger "warning: pcs -f ${cibfile} resource delete ${1}-group"
fi
shift
@@ -679,52 +667,19 @@ addnode_create_resources()
local add_vip=${1}; shift
local cibfile=$(mktemp -u)
- # mount the HA-state volume and start ganesha.nfsd on the new node
- pcs cluster cib ${cibfile}
- if [ $? -ne 0 ]; then
- logger "warning: pcs cluster cib ${cibfile} failed"
- fi
-
- pcs -f ${cibfile} resource create nfs_start-${add_node} ganesha_nfsd ha_vol_mnt=${HA_VOL_MNT}
- if [ $? -ne 0 ]; then
- logger "warning: pcs -f ${cibfile} resource create nfs_start-${add_node} ganesha_nfsd ha_vol_mnt=${HA_VOL_MNT} failed"
- fi
-
- pcs -f ${cibfile} constraint location nfs_start-${add_node} prefers ${add_node}=INFINITY
- if [ $? -ne 0 ]; then
- logger "warning: pcs -f ${cibfile} constraint location nfs_start-${add_node} prefers ${add_node}=INFINITY failed"
- fi
-
- pcs -f ${cibfile} constraint order nfs_start-${add_node} then nfs-mon-clone
- if [ $? -ne 0 ]; then
- logger "warning: pcs -f ${cibfile} constraint order nfs_start-${add_node} then nfs-mon-clone failed"
- fi
-
- pcs cluster cib-push ${cibfile}
- if [ $? -ne 0 ]; then
- logger "warning: pcs cluster cib-push ${cibfile} failed"
- fi
- rm -f ${cibfile}
-
# start HA on the new node
pcs cluster start ${add_node}
if [ $? -ne 0 ]; then
logger "warning: pcs cluster start ${add_node} failed"
fi
- pcs resource delete nfs_start-${add_node}
- if [ $? -ne 0 ]; then
- logger "warning: pcs resource delete nfs_start-${add_node} failed"
- fi
-
-
pcs cluster cib ${cibfile}
if [ $? -ne 0 ]; then
logger "warning: pcs cluster cib ${cibfile} failed"
fi
- # delete all the -cluster_ip-1 and -trigger_ip-1 resources,
- # clearing their constraints, then create them again so we can
+ # delete all the -cluster_ip-1 resources, clearing
+ # their constraints, then create them again so we can
# recompute their constraints
clear_resources ${cibfile} ${HA_SERVERS}
addnode_recreate_resources ${cibfile} ${add_node} ${add_vip}
@@ -766,31 +721,16 @@ deletenode_delete_resources()
fi
rm -f ${cibfile}
- pcs cluster cib ${cibfile}
- if [ $? -ne 0 ]; then
- logger "warning: pcs cluster cib ${cibfile} failed"
- fi
-
- pcs -f ${cibfile} resource create nfs_stop-${node} ganesha_nfsd ha_vol_mnt=${HA_VOL_MNT}
- if [ $? -ne 0 ]; then
- logger "warning: pcs -f ${cibfile} resource create nfs_stop-${node} ganesha_nfsd ha_vol_mnt=${HA_VOL_MNT} failed"
- fi
+}
- pcs -f ${cibfile} constraint location nfs_stop-${node} prefers ${node}=INFINITY
- if [ $? -ne 0 ]; then
- logger "warning: pcs -f ${cibfile} constraint location nfs_stop-${node} prefers ${node}=INFINITY failed"
- fi
- pcs cluster cib-push ${cibfile}
- if [ $? -ne 0 ]; then
- logger "warning: pcs cluster cib-push ${cibfile} failed"
- fi
- rm -f ${cibfile}
+deletenode_update_haconfig()
+{
+ local name="VIP_${1}"
+ local clean_name=${name//[-.]/_}
- pcs resource delete nfs_stop-${node}
- if [ $? -ne 0 ]; then
- logger "warning: pcs resource delete nfs_stop-${node} failed"
- fi
+ ha_servers=$(echo ${HA_SERVERS} | sed -e "s/ /,/")
+ sed -i -e "s/^HA_CLUSTER_NODES=.*$/HA_CLUSTER_NODES=\"${ha_servers// /,}\"/" -e "s/^${name}=.*$//" -e "/^$/d" ${HA_CONFDIR}/ganesha-ha.conf
}
@@ -813,7 +753,9 @@ setup_state_volume()
dirname=${1}${dname}
fi
-
+ if [ ! -d ${mnt}/nfs-ganesha/tickle_dir ]; then
+ mkdir ${mnt}/nfs-ganesha/tickle_dir
+ fi
if [ ! -d ${mnt}/nfs-ganesha/${dirname} ]; then
mkdir ${mnt}/nfs-ganesha/${dirname}
fi
@@ -825,9 +767,11 @@ setup_state_volume()
fi
if [ ! -d ${mnt}/nfs-ganesha/${dirname}/nfs/statd ]; then
mkdir ${mnt}/nfs-ganesha/${dirname}/nfs/statd
+ chown rpcuser:rpcuser ${mnt}/nfs-ganesha/${dirname}/nfs/statd
fi
if [ ! -e ${mnt}/nfs-ganesha/${dirname}/nfs/state ]; then
touch ${mnt}/nfs-ganesha/${dirname}/nfs/state
+ chown rpcuser:rpcuser ${mnt}/nfs-ganesha/${dirname}/nfs/state
fi
if [ ! -d ${mnt}/nfs-ganesha/${dirname}/nfs/ganesha/v4recov ]; then
mkdir ${mnt}/nfs-ganesha/${dirname}/nfs/ganesha/v4recov
@@ -837,15 +781,17 @@ setup_state_volume()
fi
if [ ! -d ${mnt}/nfs-ganesha/${dirname}/nfs/statd/sm ]; then
mkdir ${mnt}/nfs-ganesha/${dirname}/nfs/statd/sm
+ chown rpcuser:rpcuser ${mnt}/nfs-ganesha/${dirname}/nfs/statd/sm
fi
if [ ! -d ${mnt}/nfs-ganesha/${dirname}/nfs/statd/sm.bak ]; then
mkdir ${mnt}/nfs-ganesha/${dirname}/nfs/statd/sm.bak
+ chown rpcuser:rpcuser ${mnt}/nfs-ganesha/${dirname}/nfs/statd/sm.bak
fi
if [ ! -e ${mnt}/nfs-ganesha/${dirname}/nfs/statd/state ]; then
touch ${mnt}/nfs-ganesha/${dirname}/nfs/statd/state
fi
for server in ${HA_SERVERS} ; do
- if [ ${server} != ${dirname} ]; then
+ if [[ ${server} != ${dirname} ]]; then
ln -s ${mnt}/nfs-ganesha/${server}/nfs/ganesha ${mnt}/nfs-ganesha/${dirname}/nfs/ganesha/${server}
ln -s ${mnt}/nfs-ganesha/${server}/nfs/statd ${mnt}/nfs-ganesha/${dirname}/nfs/statd/${server}
fi
@@ -856,6 +802,214 @@ setup_state_volume()
}
+enable_pacemaker()
+{
+ while [[ ${1} ]]; do
+ if [[ "${SERVICE_MAN}" == "/bin/systemctl" ]]; then
+ ssh -oPasswordAuthentication=no -oStrictHostKeyChecking=no -i \
+${SECRET_PEM} root@${1} "${SERVICE_MAN} enable pacemaker"
+ else
+ ssh -oPasswordAuthentication=no -oStrictHostKeyChecking=no -i \
+${SECRET_PEM} root@${1} "${SERVICE_MAN} pacemaker enable"
+ fi
+ shift
+ done
+}
+
+
+addnode_state_volume()
+{
+ local newnode=${1}; shift
+ local mnt=${HA_VOL_MNT}
+ local longname=""
+ local dname=""
+ local dirname=""
+
+ longname=$(hostname)
+ dname=${longname#$(hostname -s)}
+
+ if [[ ${newnode} == *${dname} ]]; then
+ dirname=${newnode}
+ else
+ dirname=${newnode}${dname}
+ fi
+
+ if [ ! -d ${mnt}/nfs-ganesha/${dirname} ]; then
+ mkdir ${mnt}/nfs-ganesha/${dirname}
+ fi
+ if [ ! -d ${mnt}/nfs-ganesha/${dirname}/nfs ]; then
+ mkdir ${mnt}/nfs-ganesha/${dirname}/nfs
+ fi
+ if [ ! -d ${mnt}/nfs-ganesha/${dirname}/nfs/ganesha ]; then
+ mkdir ${mnt}/nfs-ganesha/${dirname}/nfs/ganesha
+ fi
+ if [ ! -d ${mnt}/nfs-ganesha/${dirname}/nfs/statd ]; then
+ mkdir ${mnt}/nfs-ganesha/${dirname}/nfs/statd
+ chown rpcuser:rpcuser ${mnt}/nfs-ganesha/${dirname}/nfs/statd
+ fi
+ if [ ! -e ${mnt}/nfs-ganesha/${dirname}/nfs/state ]; then
+ touch ${mnt}/nfs-ganesha/${dirname}/nfs/state
+ chown rpcuser:rpcuser ${mnt}/nfs-ganesha/${dirname}/nfs/state
+ fi
+ if [ ! -d ${mnt}/nfs-ganesha/${dirname}/nfs/ganesha/v4recov ]; then
+ mkdir ${mnt}/nfs-ganesha/${dirname}/nfs/ganesha/v4recov
+ fi
+ if [ ! -d ${mnt}/nfs-ganesha/${dirname}/nfs/ganesha/v4old ]; then
+ mkdir ${mnt}/nfs-ganesha/${dirname}/nfs/ganesha/v4old
+ fi
+ if [ ! -d ${mnt}/nfs-ganesha/${dirname}/nfs/statd/sm ]; then
+ mkdir ${mnt}/nfs-ganesha/${dirname}/nfs/statd/sm
+ chown rpcuser:rpcuser ${mnt}/nfs-ganesha/${dirname}/nfs/statd/sm
+ fi
+ if [ ! -d ${mnt}/nfs-ganesha/${dirname}/nfs/statd/sm.bak ]; then
+ mkdir ${mnt}/nfs-ganesha/${dirname}/nfs/statd/sm.bak
+ chown rpcuser:rpcuser ${mnt}/nfs-ganesha/${dirname}/nfs/statd/sm.bak
+ fi
+ if [ ! -e ${mnt}/nfs-ganesha/${dirname}/nfs/statd/state ]; then
+ touch ${mnt}/nfs-ganesha/${dirname}/nfs/statd/state
+ fi
+
+ for server in ${HA_SERVERS} ; do
+
+ if [[ ${server} != ${dirname} ]]; then
+ ln -s ${mnt}/nfs-ganesha/${server}/nfs/ganesha ${mnt}/nfs-ganesha/${dirname}/nfs/ganesha/${server}
+ ln -s ${mnt}/nfs-ganesha/${server}/nfs/statd ${mnt}/nfs-ganesha/${dirname}/nfs/statd/${server}
+
+ ln -s ${mnt}/nfs-ganesha/${dirname}/nfs/ganesha ${mnt}/nfs-ganesha/${server}/nfs/ganesha/${dirname}
+ ln -s ${mnt}/nfs-ganesha/${dirname}/nfs/statd ${mnt}/nfs-ganesha/${server}/nfs/statd/${dirname}
+ fi
+ done
+
+}
+
+
+delnode_state_volume()
+{
+ local delnode=${1}; shift
+ local mnt=${HA_VOL_MNT}
+ local longname=""
+ local dname=""
+ local dirname=""
+
+ longname=$(hostname)
+ dname=${longname#$(hostname -s)}
+
+ if [[ ${delnode} == *${dname} ]]; then
+ dirname=${delnode}
+ else
+ dirname=${delnode}${dname}
+ fi
+
+ rm -rf ${mnt}/nfs-ganesha/${dirname}
+
+ for server in ${HA_SERVERS} ; do
+ if [[ ${server} != ${dirname} ]]; then
+ rm -f ${mnt}/nfs-ganesha/${server}/nfs/ganesha/${dirname}
+ rm -f ${mnt}/nfs-ganesha/${server}/nfs/statd/${dirname}
+ fi
+ done
+}
+
+
+status()
+{
+ local scratch=$(mktemp)
+ local regex_str="^${1}-cluster_ip-1"
+ local healthy=0
+ local index=1
+ local nodes
+
+ # change tabs to spaces, strip leading spaces, including any
+ # new '*' at the beginning of a line introduced in pcs-0.10.x
+ pcs status | sed -e "s/\t/ /g" -e "s/^[ ]*\*//" -e "s/^[ ]*//" > ${scratch}
+
+ nodes[0]=${1}; shift
+
+ # make a regex of the configured nodes
+ # and initalize the nodes array for later
+ while [[ ${1} ]]; do
+
+ regex_str="${regex_str}|^${1}-cluster_ip-1"
+ nodes[${index}]=${1}
+ ((index++))
+ shift
+ done
+
+ # print the nodes that are expected to be online
+ grep -E "Online:" ${scratch}
+
+ echo
+
+ # print the VIPs and which node they are on
+ grep -E "${regex_str}" < ${scratch} | cut -d ' ' -f 1,4
+
+ echo
+
+ # check if the VIP and port block/unblock RAs are on the expected nodes
+ for n in ${nodes[*]}; do
+
+ grep -E -x "${n}-nfs_block \(ocf::heartbeat:portblock\): Started ${n}" > /dev/null 2>&1 ${scratch}
+ result=$?
+ ((healthy+=${result}))
+ grep -E -x "${n}-cluster_ip-1 \(ocf::heartbeat:IPaddr\): Started ${n}" > /dev/null 2>&1 ${scratch}
+ result=$?
+ ((healthy+=${result}))
+ grep -E -x "${n}-nfs_unblock \(ocf::heartbeat:portblock\): Started ${n}" > /dev/null 2>&1 ${scratch}
+ result=$?
+ ((healthy+=${result}))
+ done
+
+ grep -E "\):\ Stopped|FAILED" > /dev/null 2>&1 ${scratch}
+ result=$?
+
+ if [ ${result} -eq 0 ]; then
+ echo "Cluster HA Status: BAD"
+ elif [ ${healthy} -eq 0 ]; then
+ echo "Cluster HA Status: HEALTHY"
+ else
+ echo "Cluster HA Status: FAILOVER"
+ fi
+
+ rm -f ${scratch}
+}
+
+create_ganesha_conf_file()
+{
+ if [[ "$1" == "yes" ]];
+ then
+ if [ -e $GANESHA_CONF ];
+ then
+ rm -rf $GANESHA_CONF
+ fi
+ # The symlink /etc/ganesha/ganesha.conf need to be
+ # created using ganesha conf file mentioned in the
+ # shared storage. Every node will only have this
+ # link and actual file will stored in shared storage,
+ # so that ganesha conf editing of ganesha conf will
+ # be easy as well as it become more consistent.
+
+ ln -s $HA_CONFDIR/ganesha.conf $GANESHA_CONF
+ else
+ # Restoring previous file
+ rm -rf $GANESHA_CONF
+ cp $HA_CONFDIR/ganesha.conf $GANESHA_CONF
+ sed -r -i -e '/^%include[[:space:]]+".+\.conf"$/d' $GANESHA_CONF
+ fi
+}
+
+set_quorum_policy()
+{
+ local quorum_policy="stop"
+ local num_servers=${1}
+
+ if [ ${num_servers} -lt 3 ]; then
+ quorum_policy="ignore"
+ fi
+ pcs property set no-quorum-policy=${quorum_policy}
+ if [ $? -ne 0 ]; then
+ logger "warning: pcs property set no-quorum-policy=${quorum_policy} failed"
+ fi
+}
main()
{
@@ -865,27 +1019,29 @@ main()
usage
exit 0
fi
- if [[ ${cmd} != *status ]]; then
- HA_CONFDIR=${1%/}; shift
- local ha_conf=${HA_CONFDIR}/ganesha-ha.conf
- local node=""
- local vip=""
- # ignore any comment lines
- cfgline=$(grep ^HA_NAME= ${ha_conf})
- eval $(echo ${cfgline} | grep -F HA_NAME=)
- cfgline=$(grep ^HA_VOL_SERVER= ${ha_conf})
- eval $(echo ${cfgline} | grep -F HA_VOL_SERVER=)
- cfgline=$(grep ^HA_CLUSTER_NODES= ${ha_conf})
- eval $(echo ${cfgline} | grep -F HA_CLUSTER_NODES=)
-
- # we'll pretend that nobody ever edits /etc/os-release
- if [ -e /etc/os-release ]; then
- eval $(grep -F "REDHAT_SUPPORT_PRODUCT=" /etc/os-release)
- [ "$REDHAT_SUPPORT_PRODUCT" == "Fedora" ] && RHEL6_PCS_CNAME_OPTION=""
- fi
+ if (selinuxenabled) ;then
+ semanage boolean -m gluster_use_execmem --on
fi
+ local osid=""
+
+ osid=$(grep ^ID= /etc/os-release)
+ eval $(echo ${osid} | grep -F ID=)
+ osid=$(grep ^VERSION_ID= /etc/os-release)
+ eval $(echo ${osid} | grep -F VERSION_ID=)
+
+ HA_CONFDIR=${1%/}; shift
+ local ha_conf=${HA_CONFDIR}/ganesha-ha.conf
+ local node=""
+ local vip=""
+
+ # ignore any comment lines
+ cfgline=$(grep ^HA_NAME= ${ha_conf})
+ eval $(echo ${cfgline} | grep -F HA_NAME=)
+ cfgline=$(grep ^HA_CLUSTER_NODES= ${ha_conf})
+ eval $(echo ${cfgline} | grep -F HA_CLUSTER_NODES=)
+
case "${cmd}" in
setup | --setup)
@@ -895,17 +1051,30 @@ main()
determine_servers "setup"
- if [ "X${HA_NUM_SERVERS}X" != "X1X" ]; then
+ # Fedora 29+ and rhel/centos 8 has PCS-0.10.x
+ # default is pcs-0.10.x options but check for
+ # rhel/centos 7 (pcs-0.9.x) and adjust accordingly
+ if [[ ! ${ID} =~ {rhel,centos} ]]; then
+ if [[ ${VERSION_ID} == 7.* ]]; then
+ PCS9OR10_PCS_CNAME_OPTION="--name"
+ PCS9OR10_PCS_CLONE_OPTION="--clone"
+ fi
+ fi
+
+ if [[ "${HA_NUM_SERVERS}X" != "1X" ]]; then
+
+ determine_service_manager
setup_cluster ${HA_NAME} ${HA_NUM_SERVERS} "${HA_SERVERS}"
setup_create_resources ${HA_SERVERS}
+ setup_finalize_ha
+
setup_state_volume ${HA_SERVERS}
- setup_copy_config ${HA_SERVERS}
+ enable_pacemaker ${HA_SERVERS}
- setup_finalize
else
logger "insufficient servers for HA, aborting"
@@ -920,6 +1089,8 @@ main()
teardown_resources ${HA_SERVERS}
teardown_cluster ${HA_NAME}
+
+ cleanup_ganesha_config ${HA_CONFDIR}
;;
cleanup | --cleanup)
@@ -932,8 +1103,6 @@ main()
logger "adding ${node} with ${vip} to ${HA_NAME}"
- copy_export_config ${node} ${HA_CONFDIR}
-
determine_service_manager
manage_service "start" ${node}
@@ -946,20 +1115,26 @@ main()
fi
addnode_create_resources ${node} ${vip}
- #Subsequent add-node recreates resources for all the nodes
- #that already exist in the cluster. The nodes are picked up
- #from the entries in the ganesha-ha.conf file. Adding the
- #newly added node to the file so that the resources specfic
- #to this node is correctly recreated in the future.
- echo "VIP_$node=\"$vip\"" >> ${HA_CONFDIR}/ganesha-ha.conf
+ # Subsequent add-node recreates resources for all the nodes
+ # that already exist in the cluster. The nodes are picked up
+ # from the entries in the ganesha-ha.conf file. Adding the
+ # newly added node to the file so that the resources specfic
+ # to this node is correctly recreated in the future.
+ clean_node=${node//[-.]/_}
+ echo "VIP_${node}=\"${vip}\"" >> ${HA_CONFDIR}/ganesha-ha.conf
- NEW_NODES="$HA_CLUSTER_NODES,$node"
+ NEW_NODES="$HA_CLUSTER_NODES,${node}"
sed -i s/HA_CLUSTER_NODES.*/"HA_CLUSTER_NODES=\"$NEW_NODES\""/ \
$HA_CONFDIR/ganesha-ha.conf
- HA_SERVERS="${HA_SERVERS} ${node}"
- setup_copy_config ${HA_SERVERS}
+ addnode_state_volume ${node}
+
+ # addnode_create_resources() already appended ${node} to
+ # HA_SERVERS, so only need to increment HA_NUM_SERVERS
+ # and set quorum policy
+ HA_NUM_SERVERS=$(expr ${HA_NUM_SERVERS} + 1)
+ set_quorum_policy ${HA_NUM_SERVERS}
;;
delete | --delete)
@@ -976,20 +1151,22 @@ $HA_CONFDIR/ganesha-ha.conf
logger "warning: pcs cluster node remove ${node} failed"
fi
- ha_servers=$(echo ${HA_SERVERS} | sed -e "s/ /,/")
- sed -i "s/^HA_CLUSTER_NODES=.*$/HA_CLUSTER_NODES=\"${ha_servers// /,}\"/" ${HA_CONFDIR}/ganesha-ha.conf
-
- setup_copy_config ${HA_SERVERS}
+ deletenode_update_haconfig ${node}
- rm -rf ${HA_VOL_MNT}/nfs-ganesha/{node}
+ delnode_state_volume ${node}
determine_service_manager
manage_service "stop" ${node}
+
+ HA_NUM_SERVERS=$(expr ${HA_NUM_SERVERS} - 1)
+ set_quorum_policy ${HA_NUM_SERVERS}
;;
status | --status)
- exec pcs status
+ determine_servers "status"
+
+ status ${HA_SERVERS}
;;
refresh-config | --refresh-config)
@@ -1000,7 +1177,12 @@ $HA_CONFDIR/ganesha-ha.conf
refresh_config ${VOL} ${HA_CONFDIR} ${HA_SERVERS}
;;
- *)
+ setup-ganesha-conf-files | --setup-ganesha-conf-files)
+
+ create_ganesha_conf_file ${1}
+ ;;
+
+ *)
# setup and teardown are not intended to be used by a
# casual user
usage
@@ -1008,7 +1190,10 @@ $HA_CONFDIR/ganesha-ha.conf
;;
esac
+
+ if (selinuxenabled) ;then
+ semanage boolean -m gluster_use_execmem --off
+ fi
}
main $*
-
diff --git a/extras/ganesha/scripts/generate-epoch.py b/extras/ganesha/scripts/generate-epoch.py
new file mode 100755
index 00000000000..77af014bab9
--- /dev/null
+++ b/extras/ganesha/scripts/generate-epoch.py
@@ -0,0 +1,48 @@
+#!/usr/bin/python3
+#
+# Copyright (c) 2016 Red Hat, Inc. <http://www.redhat.com>
+# This file is part of GlusterFS.
+#
+# This file is licensed to you under your choice of the GNU Lesser
+# General Public License, version 3 or any later version (LGPLv3 or
+# later), or the GNU General Public License, version 2 (GPLv2), in all
+# cases as published by the Free Software Foundation.
+#
+# Generates unique epoch value on each gluster node to be used by
+# nfs-ganesha service on that node.
+#
+# Configure 'EPOCH_EXEC' option to this script path in
+# '/etc/sysconfig/ganesha' file used by nfs-ganesha service.
+#
+# Construct epoch as follows -
+# first 32-bit contains the now() time
+# rest 32-bit value contains the local glusterd node uuid
+
+import time
+import binascii
+
+# Calculate the now() time into a 64-bit integer value
+def epoch_now():
+ epoch_time = int(time.mktime(time.localtime())) << 32
+ return epoch_time
+
+# Read glusterd UUID and extract first 32-bit of it
+def epoch_uuid():
+ file_name = '/var/lib/glusterd/glusterd.info'
+
+ for line in open(file_name):
+ if "UUID" in line:
+ glusterd_uuid = line.split('=')[1].strip()
+
+ uuid_bin = binascii.unhexlify(glusterd_uuid.replace("-",""))
+
+ epoch_uuid = int(binascii.hexlify(uuid_bin), 32) & 0xFFFF0000
+ return epoch_uuid
+
+# Construct epoch as follows -
+# first 32-bit contains the now() time
+# rest 32-bit value contains the local glusterd node uuid
+epoch = (epoch_now() | epoch_uuid())
+print((str(epoch)))
+
+exit(0)
generated by cgit (git 2.34.1) at 2026-04-12 07:26:34 +0000