mirror of
https://github.com/ceph/ceph
synced 2024-12-23 20:03:56 +00:00
ad84ea07ca
More recent versions of ceph append a bit of whitespace to the line after the name of the /dev/rbdX device; this causes the monitor check to fail as it can't find the device name due to the whitespace. This fix excludes any characters after the /dev/rbdN match.
297 lines
8.8 KiB
Bash
297 lines
8.8 KiB
Bash
#!/bin/sh
|
|
#
|
|
# OCF resource agent for mapping and unmapping
|
|
# RADOS Block Devices (RBDs)
|
|
#
|
|
# License: GNU Lesser General Public License (LGPL) 2.1
|
|
# (c) 2012 Florian Haas, hastexo
|
|
#
|
|
|
|
# Initialization:
|
|
: ${OCF_FUNCTIONS_DIR=${OCF_ROOT}/lib/heartbeat}
|
|
. ${OCF_FUNCTIONS_DIR}/ocf-shellfuncs
|
|
|
|
# Convenience variables
|
|
# When sysconfdir isn't passed in as a configure flag,
|
|
# it's defined in terms of prefix
|
|
prefix=@prefix@
|
|
|
|
# Defaults
|
|
OCF_RESKEY_pool_default="rbd"
|
|
OCF_RESKEY_cephconf_default="@sysconfdir@/@PACKAGE_TARNAME@/@PACKAGE_TARNAME@.conf"
|
|
: ${OCF_RESKEY_pool=${OCF_RESKEY_pool_default}}
|
|
: ${OCF_RESKEY_cephconf=${OCF_RESKEY_cephconf_default}}
|
|
|
|
rbd_meta_data() {
|
|
cat <<EOF
|
|
<?xml version="1.0"?>
|
|
<!DOCTYPE resource-agent SYSTEM "ra-api-1.dtd">
|
|
<resource-agent name="rbd" version="0.1">
|
|
<version>0.1</version>
|
|
<longdesc lang="en">
|
|
Manages RADOS Block Devices (RBDs) as a highly available
|
|
resource. Maps and unmaps RBDs as needed.
|
|
</longdesc>
|
|
<shortdesc lang="en">Maps and unmaps RADOS Block Devices</shortdesc>
|
|
<parameters>
|
|
<parameter name="name" unique="0" required="1">
|
|
<longdesc lang="en">
|
|
Name of the RBD device.
|
|
</longdesc>
|
|
<shortdesc lang="en">RBD device name</shortdesc>
|
|
<content type="string"/>
|
|
</parameter>
|
|
<parameter name="pool" unique="0" required="0">
|
|
<longdesc lang="en">
|
|
Name of the RADOS pool where the RBD has been created
|
|
</longdesc>
|
|
<shortdesc lang="en">RADOS pool name</shortdesc>
|
|
<content type="string" default="${OCF_RESKEY_pool_default}"/>
|
|
</parameter>
|
|
<parameter name="snap" unique="0" required="0">
|
|
<longdesc lang="en">
|
|
Name of the device snapshot to map.
|
|
</longdesc>
|
|
<shortdesc lang="en">Snapshot name</shortdesc>
|
|
<content type="string"/>
|
|
</parameter>
|
|
<parameter name="cephconf" unique="0" required="0">
|
|
<longdesc lang="en">
|
|
Location of the Ceph configuration file
|
|
</longdesc>
|
|
<shortdesc lang="en">Ceph configuration file</shortdesc>
|
|
<content type="string" default="${OCF_RESKEY_cephconf_default}"/>
|
|
</parameter>
|
|
<parameter name="mon" unique="0" required="0">
|
|
<longdesc lang="en">
|
|
Address (or comma-separated list of addresses) of
|
|
monitor servers to connect to. Overrides values from
|
|
configuration file.
|
|
</longdesc>
|
|
<shortdesc lang="en">Monitor address(es)</shortdesc>
|
|
<content type="string"/>
|
|
</parameter>
|
|
<parameter name="user" unique="0" required="0">
|
|
<longdesc lang="en">
|
|
Username to use when mapping the device. Required
|
|
if Ceph authentication is enabled on the monitor.
|
|
</longdesc>
|
|
<shortdesc lang="en">Authentication username</shortdesc>
|
|
<content type="string"/>
|
|
</parameter>
|
|
<parameter name="secret" unique="0" required="0">
|
|
<longdesc lang="en">
|
|
File containing an authentication secret. Required
|
|
if Ceph authentication is enabled on the monitor.
|
|
</longdesc>
|
|
<shortdesc lang="en">Authentication secret file</shortdesc>
|
|
<content type="string"/>
|
|
</parameter>
|
|
</parameters>
|
|
<actions>
|
|
<action name="start" timeout="20" />
|
|
<action name="stop" timeout="20" />
|
|
<action name="monitor" timeout="20"
|
|
interval="10" depth="0" />
|
|
<action name="meta-data" timeout="5" />
|
|
<action name="validate-all" timeout="20" />
|
|
</actions>
|
|
</resource-agent>
|
|
EOF
|
|
}
|
|
|
|
rbd_usage() {
|
|
cat <<EOF
|
|
usage: $0 {start|stop|status|monitor|validate-all|meta-data}
|
|
|
|
Expects to have a fully populated OCF RA-compliant environment set.
|
|
EOF
|
|
}
|
|
|
|
# rbd command wrapper: builds an option string for invoking RBD based
|
|
# on resource parameters, and invokes rbd through ocf_run.
|
|
do_rbd() {
|
|
local rbd_options
|
|
|
|
if [ -n "${OCF_RESKEY_cephconf}" ]; then
|
|
rbd_options="$rbd_options -c ${OCF_RESKEY_cephconf}"
|
|
fi
|
|
if [ -n "${OCF_RESKEY_mon}" ]; then
|
|
rbd_options="$rbd_options -m ${OCF_RESKEY_mon}"
|
|
fi
|
|
|
|
ocf_run rbd $rbd_options $@
|
|
}
|
|
|
|
# Convenience function that uses "rbd showmapped" to retrieve the
|
|
# mapped device name from the pool, RBD name, and snapshot.
|
|
find_rbd_dev() {
|
|
local sedpat
|
|
|
|
# Example output from "rbd showmapped" (tab separated):
|
|
# id pool image snap device
|
|
# 0 rbd test - /dev/rbd0
|
|
|
|
# Build the sed pattern, substituting "-" for the snapshot name if
|
|
# it's unset
|
|
sedpat="[0-9]\+[ \t]\+${OCF_RESKEY_pool}[ \t]\+${OCF_RESKEY_name}[ \t]\+${OCF_RESKEY_snap:--}[ \t]\+\(/dev/rbd[0-9]\+\).*"
|
|
|
|
# Run rbd showmapped, filter out the header line, then try to
|
|
# extract the device name
|
|
rbd showmapped | tail -n +2 | sed -n -e "s,$sedpat,\1,p"
|
|
}
|
|
|
|
rbd_validate_all() {
|
|
# Test for configuration errors first
|
|
if [ -z "$OCF_RESKEY_name" ]; then
|
|
ocf_log err 'Required parameter "name" is unset!'
|
|
exit $OCF_ERR_CONFIGURED
|
|
fi
|
|
|
|
# Test for required binaries
|
|
check_binary rbd
|
|
|
|
return $OCF_SUCCESS
|
|
}
|
|
|
|
rbd_monitor() {
|
|
local rc
|
|
local rbd_dev
|
|
|
|
if ! [ -d /sys/bus/rbd ]; then
|
|
ocf_log debug "rbd module is not loaded"
|
|
return $OCF_NOT_RUNNING
|
|
fi
|
|
|
|
rbd_dev=`find_rbd_dev`
|
|
|
|
if [ -z "$rbd_dev" ]; then
|
|
ocf_log debug "RBD device is unmapped"
|
|
rc=$OCF_NOT_RUNNING
|
|
elif [ -b "$rbd_dev" ]; then
|
|
ocf_log debug "RBD device is mapped to $rbd_dev"
|
|
rc=$OCF_SUCCESS
|
|
else
|
|
# Device is listed, but the corresponding path is not a block
|
|
# device.
|
|
ocf_log err "$rbd_dev is not a block device!"
|
|
rc=$OCF_ERR_GENERIC
|
|
fi
|
|
|
|
return $rc
|
|
}
|
|
|
|
rbd_start() {
|
|
local rbd_map_options
|
|
local rbd_name
|
|
|
|
# if resource is already running, bail out early
|
|
if rbd_monitor; then
|
|
ocf_log info "Resource is already running"
|
|
return $OCF_SUCCESS
|
|
fi
|
|
|
|
# actually start up the resource here (make sure to immediately
|
|
# exit with an $OCF_ERR_ error code if anything goes seriously
|
|
# wrong)
|
|
|
|
if [ ! -d /sys/bus/rbd ]; then
|
|
ocf_run modprobe -v rbd || exit $OCF_ERR_INSTALLED
|
|
fi
|
|
|
|
if [ -n "${OCF_RESKEY_user}" ]; then
|
|
rbd_map_options="--id ${OCF_RESKEY_user}"
|
|
fi
|
|
if [ -n "${OCF_RESKEY_secret}" ]; then
|
|
rbd_map_options="$rbd_map_options --keyfile ${OCF_RESKEY_secret}"
|
|
fi
|
|
|
|
rbd_name="${OCF_RESKEY_pool}/${OCF_RESKEY_name}"
|
|
if [ -n "${OCF_RESKEY_snap}" ]; then
|
|
rbd_name="$rbd_name@${OCF_RESKEY_snap}"
|
|
fi
|
|
|
|
do_rbd map $rbd_name $rbd_map_options || exit $OCF_ERR_GENERIC
|
|
|
|
# After the resource has been started, check whether it started up
|
|
# correctly. If the resource starts asynchronously, the agent may
|
|
# spin on the monitor function here -- if the resource does not
|
|
# start up within the defined timeout, the cluster manager will
|
|
# consider the start action failed
|
|
while ! rbd_monitor; do
|
|
ocf_log debug "Resource has not started yet, waiting"
|
|
sleep 1
|
|
done
|
|
|
|
# only return $OCF_SUCCESS if _everything_ succeeded as expected
|
|
return $OCF_SUCCESS
|
|
}
|
|
|
|
rbd_stop() {
|
|
local rc
|
|
local rbd_dev
|
|
|
|
rbd_monitor
|
|
rc=$?
|
|
case "$rc" in
|
|
"$OCF_SUCCESS")
|
|
# Currently running. Normal, expected behavior.
|
|
ocf_log debug "Resource is currently running"
|
|
;;
|
|
"$OCF_NOT_RUNNING")
|
|
# Currently not running. Nothing to do.
|
|
ocf_log info "Resource is already stopped"
|
|
return $OCF_SUCCESS
|
|
;;
|
|
esac
|
|
|
|
# actually shut down the resource here (make sure to immediately
|
|
# exit with an $OCF_ERR_ error code if anything goes seriously
|
|
# wrong)
|
|
rbd_dev=`find_rbd_dev`
|
|
do_rbd unmap $rbd_dev || exit $OCF_ERR_GENERIC
|
|
|
|
# After the resource has been stopped, check whether it shut down
|
|
# correctly. If the resource stops asynchronously, the agent may
|
|
# spin on the monitor function here -- if the resource does not
|
|
# shut down within the defined timeout, the cluster manager will
|
|
# consider the stop action failed
|
|
while rbd_monitor; do
|
|
ocf_log debug "Resource has not stopped yet, waiting"
|
|
sleep 1
|
|
done
|
|
|
|
# only return $OCF_SUCCESS if _everything_ succeeded as expected
|
|
return $OCF_SUCCESS
|
|
|
|
}
|
|
|
|
# Make sure meta-data and usage always succeed
|
|
case $__OCF_ACTION in
|
|
meta-data) rbd_meta_data
|
|
exit $OCF_SUCCESS
|
|
;;
|
|
usage|help) rbd_usage
|
|
exit $OCF_SUCCESS
|
|
;;
|
|
esac
|
|
|
|
# Anything other than meta-data and usage must pass validation
|
|
rbd_validate_all || exit $?
|
|
|
|
# Translate each action into the appropriate function call
|
|
case $__OCF_ACTION in
|
|
start) rbd_start;;
|
|
stop) rbd_stop;;
|
|
status|monitor) rbd_monitor;;
|
|
validate-all) ;;
|
|
*) rbd_usage
|
|
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"
|
|
exit $rc
|