With Type=notify services, EXTEND_TIMEOUT_USEC= messages will delay any startup/
runtime/shutdown timeouts.
A service that hasn't timed out, i.e, start time < TimeStartSec,
runtime < RuntimeMaxSec and stop time < TimeoutStopSec, may by sending
EXTEND_TIMEOUT_USEC=, allow the service to continue beyond the limit for
the execution phase (i.e TimeStartSec, RunTimeMaxSec and TimeoutStopSec).
EXTEND_TIMEOUT_USEC= must continue to be sent (in the same way as
WATCHDOG=1) within the time interval specified to continue to reprevent
the timeout from occuring.
Watchdog timeouts are also extended if a EXTEND_TIMEOUT_USEC is greater
than the remaining time on the watchdog counter.
Fixes #5868.
Example : <literal>WATCHDOG_USEC=20000000</literal></para></listitem>
</varlistentry>
+ <varlistentry>
+ <term>EXTEND_TIMEOUT_USEC=…</term>
+
+ <listitem><para>Tells the service manager to extend the startup, runtime or shutdown service timeout
+ corresponding the current state. The value specified is a time in microseconds during which the service must
+ send a new message. A service timeout will occur if the message isn't received, but only if the runtime of the
+ current state is beyond the original maximium times of <varname>TimeoutStartSec=</varname>, <varname>RuntimeMaxSec=</varname>,
+ and <varname>TimeoutStopSec=</varname>.
+ See <citerefentry><refentrytitle>elogind.service</refentrytitle><manvolnum>5</manvolnum></citerefentry>
+ for effects on the service timeouts.</para></listitem>
+ </varlistentry>
+
<varlistentry>
<term>FDSTORE=1</term>
--- /dev/null
+#!/bin/bash
+set -v -x
+
+rm -f /test.log
+
+TL=/test.log.XXXXXXXX
+
+function wait_for()
+{
+ service=${1}
+ result=${2:-success}
+ time=${3:-45}
+
+ while [[ ! -f /${service}.terminated && ! -f /${service}.success && $time -gt 0 ]]
+ do
+ sleep 1
+ time=$(( $time - 1 ))
+ done
+
+ if [[ ! -f /${service}.${result} ]]
+ then
+ journalctl -u testsuite-${service/_/-}.service >> "${TL}"
+ fi
+}
+
+# This checks all stages, start, runtime and stop, can be extended by
+# EXTEND_TIMEOUT_USEC
+
+wait_for success_all
+
+# These check that EXTEND_TIMEOUT_USEC that occurs at greater than the
+# extend timeout interval but less then the stage limit (TimeoutStartSec,
+# RuntimeMaxSec, TimeoutStopSec) still succeed.
+
+wait_for success_start
+wait_for success_runtime
+wait_for success_stop
+
+# These ensure that EXTEND_TIMEOUT_USEC will still timeout in the
+# approprate stage, after the stage limit, when the EXTEND_TIMEOUT_USEC
+# message isn't sent within the extend timeout interval.
+
+wait_for fail_start startfail
+wait_for fail_stop stopfail
+wait_for fail_runtime runtimefail
+
+if [[ -f "${TL}" ]]
+then
+ # no mv
+ cp "${TL}" /test.log
+ exit 1
+else
+ touch /testok
+ exit 0
+fi
--- /dev/null
+#!/bin/bash
+set -x
+set -e
+set -o pipefail
+
+# sleep interval (seconds)
+sleep_interval=1
+# extend_timeout_interval second(s)
+extend_timeout_interval=1
+# number of sleep_intervals before READY=1
+start_intervals=10
+# number of sleep_intervals before exiting
+stop_intervals=10
+# run intervals, number of sleep_intervals to run
+run_intervals=7
+# service name
+SERVICE=unknown
+
+while [ $# -gt 0 ];
+do
+ eval ${1%=*}=${1#*=}
+ shift
+done
+
+# We convert to usec
+extend_timeout_interval=$(( $extend_timeout_interval * 1000000 ))
+
+trap "{ touch /${SERVICE}.terminated; exit 1; }" SIGTERM SIGABRT
+
+rm -f /${SERVICE}.*
+touch /${SERVICE}.startfail
+
+elogind-notify EXTEND_TIMEOUT_USEC=$extend_timeout_interval
+while [ $start_intervals -gt 0 ]
+do
+ sleep $sleep_interval
+ start_intervals=$(( $start_intervals - 1 ))
+ elogind-notify EXTEND_TIMEOUT_USEC=$extend_timeout_interval
+done
+
+elogind-notify --ready --status="Waiting for your request"
+
+touch /${SERVICE}.runtimefail
+rm /${SERVICE}.startfail
+
+elogind-notify EXTEND_TIMEOUT_USEC=$extend_timeout_interval
+while [ $run_intervals -gt 0 ]
+do
+ sleep $sleep_interval
+ run_intervals=$(( $run_intervals - 1 ))
+ elogind-notify EXTEND_TIMEOUT_USEC=$extend_timeout_interval
+done
+
+elogind-notify STOPPING=1
+
+touch /${SERVICE}.stopfail
+rm /${SERVICE}.runtimefail
+
+elogind-notify EXTEND_TIMEOUT_USEC=$extend_timeout_interval
+while [ $stop_intervals -gt 0 ]
+do
+ sleep $sleep_interval
+ stop_intervals=$(( $stop_intervals - 1 ))
+ elogind-notify EXTEND_TIMEOUT_USEC=$extend_timeout_interval
+done
+
+touch /${SERVICE}.success
+rm /${SERVICE}.stopfail
+
+exit 0
--- /dev/null
+
+[Unit]
+Description=Testsuite: Fail Runtime (EXTEND_TIMEOUT_USEC Didn't occur in sufficient time after RuntimeSecMax.)
+
+[Service]
+
+# EXTEND_TIMEOUT_USEC on runtime start (0) and 7 seconds after. Systemd will expect one at 7+5 (extend_timeout_interval)
+# seconds this won't happen until 7 + 7 (sleep interval) seconds. Therefore timeout at 12 seconds.
+Type=notify
+TimeoutStartSec=4
+TimeoutStopSec=4
+RuntimeMaxSec=10
+ExecStart=/extend_timeout_test_service.sh SERVICE=fail_runtime extend_timeout_interval=5 sleep_interval=7 start_intervals=0 run_intervals=2 stop_intervals=0
--- /dev/null
+
+[Unit]
+Description=Testsuite: Fail Start (EXTEND_TIMEOUT_USEC Didn't occur in sufficient time after TimeoutStartSec.)
+
+[Service]
+
+# EXTEND_TIMEOUT_USEC on startup and 7 seconds from start. Systemd will expect one at 7+5 (extend_timeout_interval)
+# seconds this won't happen until 7 + 7 (sleep interval) seconds. Therefore timeout at 12 seconds.
+Type=notify
+TimeoutStartSec=10
+TimeoutStopSec=4
+RuntimeMaxSec=4
+ExecStart=/extend_timeout_test_service.sh SERVICE=fail_start extend_timeout_interval=5 sleep_interval=7 start_intervals=2 run_intervals=0 stop_intervals=0
--- /dev/null
+
+[Unit]
+Description=Testsuite: Fail Stop (EXTEND_TIMEOUT_USEC Didn't occur in sufficient time after TimeoutStopSec.)
+
+[Service]
+
+# EXTEND_TIMEOUT_USEC on stop (0) and 7 seconds after. Systemd will expect one at 7+5 (extend_timeout_interval)
+# seconds this won't happen until 7 + 7 (sleep interval) seconds. Therefore timeout at 12 seconds.
+Type=notify
+TimeoutStartSec=4
+TimeoutStopSec=10
+RuntimeMaxSec=4
+ExecStart=/extend_timeout_test_service.sh SERVICE=fail_stop extend_timeout_interval=5 sleep_interval=7 start_intervals=0 run_intervals=0 stop_intervals=2
+# Due to 6041a7ee2c1bbff6301082f192fc1b0882400d42 SIGTERM isn't sent as the service shuts down with STOPPING=1
+# This file makes the test assess.sh quicker by notifing it that this test has finished.
+ExecStopPost=/bin/bash -c '[[ $SERVICE_RESULT == timeout && $EXIT_CODE == killed ]] && touch /fail_runtime.terminated'
--- /dev/null
+
+[Unit]
+Description=Testsuite: EXTEND_TIMEOUT_USEC Success - extend timeout on all services
+
+[Service]
+
+# Normal success - startup / runtime / shutdown all take 8 seconds which is within the EXTEND_TIMEOUT_USEC=4 seconds interval
+# runtime is 8+8+8 seconds. so we are relying on the EXTEND_TIMEOUT_USEC to exceed all stages, Start, Runtime and Stop.
+# success occurs after 24 seconds
+Type=notify
+TimeoutStartSec=4
+TimeoutStopSec=4
+RuntimeMaxSec=4
+ExecStart=/extend_timeout_test_service.sh SERVICE=success_all extend_timeout_interval=4 sleep_interval=2 start_intervals=3 run_intervals=3 stop_intervals=3
--- /dev/null
+
+[Unit]
+Description=Testsuite: Success Runtime (EXTEND_TIMEOUT_USEC > WATCHDOG_USEC however < RuntimeMaxSec)
+
+[Service]
+
+# EXTEND_TIMEOUT_USEC=4 second once during runtime, but sleep for 6 seconds.
+# Runtime is 6 seconds and < RuntimeMaxSec so still successful.
+Type=notify
+TimeoutStartSec=4
+TimeoutStopSec=4
+RuntimeMaxSec=8
+ExecStart=/extend_timeout_test_service.sh SERVICE=success_runtime extend_timeout_interval=4 sleep_interval=6 start_intervals=0 run_intervals=1 stop_intervals=0
--- /dev/null
+
+[Unit]
+Description=Testsuite: Success Start (EXTEND_TIMEOUT_USEC > WATCHDOG_USEC however < TimeoutStartSec)
+
+[Service]
+
+# EXTEND_TIMEOUT_USEC=4 second interval once at startup, but sleep 6 seconds.
+# Therefore startup is 6 seconds and < TimeoutStartSec so still successful.
+Type=notify
+TimeoutStartSec=8
+TimeoutStopSec=4
+RuntimeMaxSec=4
+ExecStart=/extend_timeout_test_service.sh SERVICE=success_start extend_timeout_interval=4 sleep_interval=6 start_intervals=1 run_intervals=0 stop_intervals=0
--- /dev/null
+
+[Unit]
+Description=Testsuite: Success Stop (EXTEND_TIMEOUT_USEC > WATCHDOG_USEC however < TimeoutStopSec)
+
+[Service]
+
+# EXTEND_TIMEOUT_USEC=4 seconds once during shutdown, but sleep for 6 seconds.
+# Therefore stop time is 6 seconds and < TimeoutStopSec so still successful.
+Type=notify
+TimeoutStartSec=4
+TimeoutStopSec=8
+RuntimeMaxSec=4
+ExecStart=/extend_timeout_test_service.sh SERVICE=success_stop extend_timeout_interval=4 sleep_interval=6 start_intervals=0 run_intervals=0 stop_intervals=1
--- /dev/null
+
+[Unit]
+Description=Testsuite: Assess all other testsuite-*.services worked as expected
+
+Wants=testsuite-success-all.service
+Wants=testsuite-success-start.service
+Wants=testsuite-success-runtime.service
+Wants=testsuite-success-stop.service
+Wants=testsuite-fail-start.service
+Wants=testsuite-fail-stop.service
+Wants=testsuite-fail-runtime.service
+StopWhenUnneeded=yes
+
+[Service]
+
+Type=simple
+ExecStartPre=/assess.sh
+ExecStart=/bin/true
~ianmdlvl / elogind.git / commitdiff