##############################################################
-# CTDB_SCRIPT_DEBUGLEVEL can be overwritten by setting it in a
-# configuration file.
-debug ()
-{
- if [ "${CTDB_SCRIPT_DEBUGLEVEL:-2}" -ge 4 ] ; then
- # If there are arguments then echo them. Otherwise expect to
- # use stdin, which allows us to pass lots of debug using a
- # here document.
- if [ -n "$1" ] ; then
- echo "DEBUG: $*"
- else
- sed -e 's@^@DEBUG: @'
- fi
- else
- if [ -z "$1" ] ; then
- cat >/dev/null
- fi
- fi
-}
-
die ()
{
_msg="$1"
######################################################
# Cached retrieval of PNN from local node. This never changes so why
# open a client connection to the server each time this is needed?
-# This sets $pnn - this avoid an unnecessary subprocess.
ctdb_get_pnn ()
{
_pnn_file="${CTDB_SCRIPT_VARDIR}/my-pnn"
if [ ! -f "$_pnn_file" ] ; then
- $CTDB pnn | sed -e 's@.*:@@' >"$_pnn_file"
+ $CTDB pnn >"$_pnn_file"
fi
- read pnn <"$_pnn_file"
+ cat "$_pnn_file"
}
# Cached retrieval of private IP address from local node. This never
-# changes. Sets $ip_address to avoid an unnecessary subprocess.
+# changes.
ctdb_get_ip_address ()
{
_ip_addr_file="${CTDB_SCRIPT_VARDIR}/my-ip-address"
# ip_address is used by caller
# shellcheck disable=SC2034
- read ip_address <"$_ip_addr_file"
+ cat "$_ip_addr_file"
}
######################################################
# usage: ctdb_check_tcp_ports <ports...>
######################################################
-# This flag file is created when a service is initially started. It
-# is deleted the first time TCP port checks for that service succeed.
-# Until then ctdb_check_tcp_ports() prints a more subtle "error"
-# message if a port check fails.
-_ctdb_check_tcp_common ()
-{
- assert_service_name
- _d="${CTDB_SCRIPT_VARDIR}/failcount"
- _ctdb_service_started_file="${_d}/${service_name}.started"
-}
-
-ctdb_check_tcp_init ()
-{
- _ctdb_check_tcp_common
- mkdir -p "${_ctdb_service_started_file%/*}" # dirname
- touch "$_ctdb_service_started_file"
-}
-
# Check whether something is listening on all of the given TCP ports
# using the "ctdb checktcpport" command.
ctdb_check_tcp_ports()
{
- if [ -z "$1" ] ; then
- echo "INTERNAL ERROR: ctdb_check_tcp_ports - no ports specified"
- exit 1
- fi
-
- for _p ; do # process each function argument (port)
- _cmd="$CTDB checktcpport $_p"
- _out=$($_cmd 2>&1)
- _ret=$?
- case "$_ret" in
- 0)
- _ctdb_check_tcp_common
- if [ ! -f "$_ctdb_service_started_file" ] ; then
- echo "ERROR: $service_name tcp port $_p is not responding"
- debug "\"ctdb checktcpport $_p\" was able to bind to port"
- else
- echo "INFO: $service_name tcp port $_p is not responding"
- fi
+ if [ -z "$1" ] ; then
+ echo "INTERNAL ERROR: ctdb_check_tcp_ports - no ports specified"
+ exit 1
+ fi
- return 1
- ;;
- 98)
- # Couldn't bind, something already listening, next port...
- continue
- ;;
- *)
- echo "ERROR: unexpected error running \"ctdb checktcpport\""
- debug <<EOF
-$CTDB checktcpport (exited with $_ret) with output:
-$_out"
-EOF
- return $_ret
- esac
- done
+ for _p ; do # process each function argument (port)
+ _cmd="$CTDB checktcpport $_p"
+ _out=$($_cmd 2>&1)
+ _ret=$?
+ case "$_ret" in
+ 0)
+ echo "$service_name not listening on TCP port $_p"
+ return 1
+ ;;
+ 98)
+ # Couldn't bind, something already listening, next port
+ continue
+ ;;
+ *)
+ echo "unexpected error (${_ret}) running \"${_cmd}\""
+ if [ -n "$_out" ] ; then
+ echo "$_out"
+ fi
+ return $_ret
+ ;;
+ esac
+ done
- # All ports listening
- _ctdb_check_tcp_common
- rm -f "$_ctdb_service_started_file"
- return 0
+ # All ports listening
+ return 0
}
######################################################
fi
}
-######################################################
-# check a command returns zero status
-# usage: ctdb_check_command <command>
-######################################################
-ctdb_check_command ()
-{
- _out=$("$@" 2>&1) || {
- echo "ERROR: $* returned error"
- echo "$_out" | debug
- exit 1
- }
-}
-
################################################
# kill off any TCP connections with the given IP
################################################
139|445) __oneway=true ;;
esac
- echo "Killing TCP connection $_src $_dst"
_connections="${_connections}${_nl}${_src} ${_dst}"
if ! $__oneway ; then
_connections="${_connections}${_nl}${_dst} ${_src}"
return
}
- _remaining=$(get_tcp_connections_for_ip "$_ip" | wc -l)
-
- if [ "$_remaining" -eq 0 ] ; then
- echo "Killed $_killcount TCP connections to released IP $_ip"
- return
+ _connections=$(get_tcp_connections_for_ip "$_ip")
+ if [ -z "$_connections" ] ; then
+ _remaining=0
+ else
+ _remaining=$(echo "$_connections" | wc -l)
fi
- _t="${_remaining}/${_killcount}"
- echo "Failed to kill TCP connections for IP $_ip (${_t} remaining)"
+ _actually_killed=$((_killcount - _remaining))
+
+ _t="${_actually_killed}/${_killcount}"
+ echo "Killed ${_t} TCP connections to released IP $_ip"
+
+ if [ -n "$_connections" ] ; then
+ echo "Remaining connections:"
+ echo "$_connections" | sed -e 's|^| |'
+ fi
}
}
{
_ip="$1"
- get_tcp_connections_for_ip "$_ip" |
- {
- _failed=false
+ # Get connections, both directions
+ _conns=$(get_tcp_connections_for_ip "$_ip" | \
+ awk '{ print $1, $2 ; print $2, $1 }')
- while read dest src; do
- echo "Tickle TCP connection $src $dest"
- $CTDB tickle "$src" "$dest" >/dev/null 2>&1 || _failed=true
- echo "Tickle TCP connection $dest $src"
- $CTDB tickle "$dest" "$src" >/dev/null 2>&1 || _failed=true
- done
-
- if $_failed ; then
- echo "Failed to send tickle control"
- fi
- }
+ echo "$_conns" | awk '{ print "Tickle TCP connection", $1, $2 }'
+ echo "$_conns" | ctdb tickle
}
get_tcp_connections_for_ip ()
*:*) _bits=128 ;;
*) _bits=32 ;;
esac
-
ip addr show to "${_addr}/${_bits}" 2>/dev/null | \
- awk 'NR == 1 { iface = $2; sub(":$", "", iface) ; \
- sub("@.*", "", iface) } \
- $1 ~ /inet/ { mask = $2; sub(".*/", "", mask); \
+ awk 'NR == 1 { iface = $2; sub(":$", "", iface) ;
+ sub("@.*", "", iface) }
+ $1 ~ /inet/ { mask = $2; sub(".*/", "", mask);
print mask, iface }'
}
# Extract the full interface description to see if it is a VLAN
_t=$(echo "$_iface_info" |
- awk 'NR == 1 { iface = $2; sub(":$", "", iface) ; \
+ awk 'NR == 1 { iface = $2; sub(":$", "", iface) ;
print iface }')
case "$_t" in
*@*)
_counter_file="${CTDB_SCRIPT_VARDIR}/failcount/${_service_name}"
mkdir -p "${_counter_file%/*}" # dirname
}
+# Some code passes an argument
+# shellcheck disable=SC2120
ctdb_counter_init () {
_ctdb_counter_common "$1"
ctdb_counter_incr () {
_ctdb_counter_common "$1"
- # unary counting!
- echo -n 1 >> "$_counter_file"
+ # unary counting using newlines!
+ echo >>"$_counter_file"
}
ctdb_counter_get () {
_ctdb_counter_common "$1"
ctdb_setup_service_state_dir ()
{
- service_state_dir="${CTDB_SCRIPT_VARDIR}/service_state/${1:-${service_name}}"
- mkdir -p "$service_state_dir" || {
- echo "Error creating state dir \"$service_state_dir\""
- exit 1
- }
-}
-
-########################################################
-# Managed status history, for auto-start/stop
-
-_ctdb_managed_common ()
-{
- _ctdb_managed_file="${CTDB_SCRIPT_VARDIR}/managed_history/${service_name}"
-}
+ _s="${1:-${service_name}}"
-ctdb_service_managed ()
-{
- _ctdb_managed_common
- mkdir -p "${_ctdb_managed_file%/*}" # dirname
- touch "$_ctdb_managed_file"
-}
+ _service_state_dir="${CTDB_SCRIPT_VARDIR}/service_state/${_s}"
+ mkdir -p "$_service_state_dir" ||
+ die "Error creating state dir \"${_service_state_dir}\""
-ctdb_service_unmanaged ()
-{
- _ctdb_managed_common
- rm -f "$_ctdb_managed_file"
-}
-
-is_ctdb_previously_managed_service ()
-{
- _ctdb_managed_common
- [ -f "$_ctdb_managed_file" ]
+ echo "$_service_state_dir"
}
##################################################################
echo "Reconfiguring service \"${service_name}\"..."
ctdb_service_unset_reconfigure
service_reconfigure || return $?
+ # Intentionally have this use $service_name as default
+ # shellcheck disable=SC2119
ctdb_counter_init
}
:
}
-ctdb_reconfigure_take_lock ()
-{
- _ctdb_service_reconfigure_common
- _lock="${_d}/reconfigure_lock"
- mkdir -p "${_lock%/*}" # dirname
- touch "$_lock"
-
- (
- flock 0
- # This is overkill but will work if we need to extend this to
- # allow certain events to run multiple times in parallel
- # (e.g. takeip) and write multiple PIDs to the file.
- read _locker_event
- if [ -n "$_locker_event" ] ; then
- while read _pid ; do
- if [ -n "$_pid" -a "$_pid" != $$ ] && \
- kill -0 "$_pid" 2>/dev/null ; then
- exit 1
- fi
- done
- fi
-
- printf "%s\n%s\n" "$event_name" $$ >"$_lock"
- exit 0
- ) <"$_lock"
-}
-
-ctdb_reconfigure_release_lock ()
-{
- _ctdb_service_reconfigure_common
- _lock="${_d}/reconfigure_lock"
-
- rm -f "$_lock"
-}
-
-ctdb_replay_monitor_status ()
-{
- echo "Replaying previous status for this script due to reconfigure..."
- # Leading separator ('|') is missing in some versions...
- _out=$($CTDB scriptstatus -X | grep -E "^\|?monitor\|${script_name}\|")
- # Output looks like this:
- # |monitor|60.nfs|1|ERROR|1314764004.030861|1314764004.035514|foo bar|
- # This is the cheapest way of getting fields in the middle.
- # Intentional word splitting here
- # shellcheck disable=SC2046,2086
- set -- $(IFS="|" ; echo $_out)
- _code="$3"
- _status="$4"
- # The error output field can include colons so we'll try to
- # preserve them. The weak checking at the beginning tries to make
- # this work for both broken (no leading '|') and fixed output.
- _out="${_out%|}"
- _err_out="${_out#*monitor|${script_name}|*|*|*|*|}"
- case "$_status" in
- OK) : ;; # Do nothing special.
- TIMEDOUT)
- # Recast this as an error, since we can't exit with the
- # correct negative number.
- _code=1
- _err_out="[Replay of TIMEDOUT scriptstatus - note incorrect return code.] ${_err_out}"
- ;;
- DISABLED)
- # Recast this as an OK, since we can't exit with the
- # correct negative number.
- _code=0
- _err_out="[Replay of DISABLED scriptstatus - note incorrect return code.] ${_err_out}"
- ;;
- *) : ;; # Must be ERROR, do nothing special.
- esac
- if [ -n "$_err_out" ] ; then
- echo "$_err_out"
- fi
- exit $_code
-}
-
-ctdb_service_check_reconfigure ()
-{
- assert_service_name
-
- # We only care about some events in this function. For others we
- # return now.
- case "$event_name" in
- monitor|ipreallocated|reconfigure) : ;;
- *) return 0 ;;
- esac
-
- if ctdb_reconfigure_take_lock ; then
- # No events covered by this function are running, so proceed
- # with gay abandon.
- case "$event_name" in
- reconfigure)
- (ctdb_service_reconfigure)
- exit $?
- ;;
- ipreallocated)
- if ctdb_service_needs_reconfigure ; then
- ctdb_service_reconfigure
- fi
- ;;
- esac
-
- ctdb_reconfigure_release_lock
- else
- # Somebody else is running an event we don't want to collide
- # with. We proceed with caution.
- case "$event_name" in
- reconfigure)
- # Tell whoever called us to retry.
- exit 2
- ;;
- ipreallocated)
- # Defer any scheduled reconfigure and just run the
- # rest of the ipreallocated event, as per the
- # eventscript. There's an assumption here that the
- # event doesn't depend on any scheduled reconfigure.
- # This is true in the current code.
- return 0
- ;;
- monitor)
- # There is most likely a reconfigure in progress so
- # the service is possibly unstable. As above, we
- # defer any scheduled reconfigured. We also replay
- # the previous monitor status since that's the best
- # information we have.
- ctdb_replay_monitor_status
- ;;
- esac
- fi
-}
-
##################################################################
-# Does CTDB manage this service? - and associated auto-start/stop
+# Does CTDB manage this service?
ctdb_compat_managed_service ()
{
[ "${t#* ${service_name} }" != "${t}" ]
}
-ctdb_start_stop_service ()
-{
- assert_service_name
-
- # Allow service-start/service-stop pseudo-events to start/stop
- # services when we're not auto-starting/stopping and we're not
- # monitoring.
- case "$event_name" in
- service-start)
- if is_ctdb_managed_service ; then
- die 'service-start event not permitted when service is managed'
- fi
- if [ "$CTDB_SERVICE_AUTOSTARTSTOP" = "yes" ] ; then
- die 'service-start event not permitted with CTDB_SERVICE_AUTOSTARTSTOP=yes'
- fi
- ctdb_service_start
- exit $?
- ;;
- service-stop)
- if is_ctdb_managed_service ; then
- die 'service-stop event not permitted when service is managed'
- fi
- if [ "$CTDB_SERVICE_AUTOSTARTSTOP" = "yes" ] ; then
- die 'service-stop event not permitted with CTDB_SERVICE_AUTOSTARTSTOP=yes'
- fi
- ctdb_service_stop
- exit $?
- ;;
- esac
-
- # Do nothing unless configured to...
- [ "$CTDB_SERVICE_AUTOSTARTSTOP" = "yes" ] || return 0
-
- [ "$event_name" = "monitor" ] || return 0
-
- if is_ctdb_managed_service ; then
- if ! is_ctdb_previously_managed_service ; then
- echo "Starting service \"$service_name\" - now managed"
- background_with_logging ctdb_service_start
- exit $?
- fi
- else
- if is_ctdb_previously_managed_service ; then
- echo "Stopping service \"$service_name\" - no longer managed"
- background_with_logging ctdb_service_stop
- exit $?
- fi
- fi
-}
-
-ctdb_service_start ()
-{
- # The service is marked managed if we've ever tried to start it.
- ctdb_service_managed
-
- service_start || return $?
-
- ctdb_counter_init
- ctdb_check_tcp_init
-}
-
-ctdb_service_stop ()
-{
- ctdb_service_unmanaged
- service_stop
-}
-
# Default service_start() and service_stop() functions.
# These may be overridden in an eventscript.
}
# AIX (and perhaps others?) doesn't have mktemp
+# type is commonly supported and more portable than which(1)
+# shellcheck disable=SC2039
if ! type mktemp >/dev/null 2>&1 ; then
mktemp ()
{
nfs_callout_init ()
{
+ _state_dir="$1"
+
if [ -z "$CTDB_NFS_CALLOUT" ] ; then
CTDB_NFS_CALLOUT="${CTDB_BASE}/nfs-linux-kernel-callout"
fi
export CTDB_NFS_CALLOUT
# If the callout wants to use this then it must create it
- export CTDB_NFS_CALLOUT_STATE_DIR="${service_state_dir}/callout-state"
+ export CTDB_NFS_CALLOUT_STATE_DIR="${_state_dir}/callout-state"
# Export, if set, for use by clustered NFS callouts
if [ -n "$CTDB_NFS_STATE_FS_TYPE" ] ; then
export CTDB_NFS_STATE_MNT
fi
- nfs_callout_cache="${service_state_dir}/nfs_callout_cache"
+ nfs_callout_cache="${_state_dir}/nfs_callout_cache"
nfs_callout_cache_callout="${nfs_callout_cache}/CTDB_NFS_CALLOUT"
nfs_callout_cache_ops="${nfs_callout_cache}/ops"
}
tickledir="${CTDB_SCRIPT_VARDIR}/tickles"
mkdir -p "$tickledir"
- ctdb_get_pnn
-
# What public IPs do I hold?
- _ips=$($CTDB -X ip | awk -F'|' -v pnn="$pnn" '$3 == pnn {print $2}')
+ _pnn=$(ctdb_get_pnn)
+ _ips=$($CTDB -X ip | awk -F'|' -v pnn="$_pnn" '$3 == pnn {print $2}')
# IPs and port as ss filters
_ip_filter=""
sort >"$_my_tickles"
# Add tickles for connections that we haven't already got tickles for
- comm -23 "$_my_connections" "$_my_tickles" |
- while read _src _dst ; do
- $CTDB addtickle "$_src" "$_dst"
- done
+ comm -23 "$_my_connections" "$_my_tickles" | \
+ $CTDB addtickle
# Remove tickles for connections that are no longer there
- comm -13 "$_my_connections" "$_my_tickles" |
- while read _src _dst ; do
- $CTDB deltickle "$_src" "$_dst"
- done
+ comm -13 "$_my_connections" "$_my_tickles" | \
+ $CTDB deltickle
rm -f "$_my_connections" "$_my_tickles"
# Remove stale files from killed scripts
- find "$tickledir" -type f -mmin +10 | xargs -r rm
+ # Files can't have spaces in name, more portable than -print0/-0
+ # shellcheck disable=SC2038
+ (cd "$tickledir" && find . -type f -mmin +10 | xargs -r rm)
}
########################################################
}
script_name="${0##*/}" # basename
-event_name="$1"