|
|
Yes bacula and as such bareos exits with an exit code that represents the
signal received. Its seems systemd has a other idea of how signal handling
should work and you should catch your signal do post-processing and then
remove the signal handler and then send the signal once again and let the
the default handler cleanup your running process.
For now I think adding the also clean exit codes like 15 makes sense.
Write a patch and attach it to the bug and then we can merge it into master
and see how it works out there. |
|
|
|
Fix-386_make-systemd-aware-of-normal-exit-status.patch (1,999 bytes)
From 2e62972276840603cf1b194e6097a30e07a507d6 Mon Sep 17 00:00:00 2001
From: Bruno Friedmann <bruno@ioda-net.ch>
Date: Sun, 4 Jan 2015 09:47:30 +0100
Subject: [PATCH 1/1] Make systemd aware of normal daemon exit status, so
service are in success state after stop or restart
---
platforms/systemd/bareos-dir.service.in | 1 +
platforms/systemd/bareos-fd.service.in | 1 +
platforms/systemd/bareos-sd.service.in | 1 +
3 files changed, 3 insertions(+)
diff --git a/platforms/systemd/bareos-dir.service.in b/platforms/systemd/bareos-dir.service.in
index 0ebed66..c94d17c 100644
--- a/platforms/systemd/bareos-dir.service.in
+++ b/platforms/systemd/bareos-dir.service.in
@@ -34,6 +34,7 @@ PIDFile=@piddir@/bareos-dir.@dir_port@.pid
# StandardOutput=syslog
ExecStartPre=@sbindir@/bareos-dir -c @confdir@/bareos-dir.conf -t -f
ExecStart=@sbindir@/bareos-dir -c @confdir@/bareos-dir.conf
+SuccessExitStatus=0 1 15
# This daemon should be able to reload the conf file
#ExecReload=/sbin/killproc -p @piddir@/bareos-dir.pid -HUP @sbindir@/bareos-dir
#Restart=on-failure
diff --git a/platforms/systemd/bareos-fd.service.in b/platforms/systemd/bareos-fd.service.in
index 1cde296..5713d6c 100644
--- a/platforms/systemd/bareos-fd.service.in
+++ b/platforms/systemd/bareos-fd.service.in
@@ -28,6 +28,7 @@ WorkingDirectory=@working_dir@
PIDFile=@piddir@/bareos-fd.@fd_port@.pid
StandardOutput=syslog
ExecStart=@sbindir@/bareos-fd -c @confdir@/bareos-fd.conf
+SuccessExitStatus=0 15
Restart=on-failure
# IOSchedulingClass=idle
diff --git a/platforms/systemd/bareos-sd.service.in b/platforms/systemd/bareos-sd.service.in
index d6d8ab9..fd88d75 100644
--- a/platforms/systemd/bareos-sd.service.in
+++ b/platforms/systemd/bareos-sd.service.in
@@ -29,6 +29,7 @@ PIDFile=@piddir@/bareos-sd.@sd_port@.pid
# EnvironmentFile=-/etc/sysconfig/bareos-sd
StandardOutput=syslog
ExecStart=@sbindir@/bareos-sd -c @confdir@/bareos-sd.conf
+SuccessExitStatus=0 15
Restart=on-failure
[Install]
--
1.8.4.5
|
|
|
|
Find attached patch against current master branch
(Should be easy to backport)
I found that bareos-dir need also signal 1 to have success when doing a restart. |
|
|
|
fixed in master with changeset c862034f |
|
|
|
Fix committed to bareos2015 bareos-14.2 branch with changesetid 4471. |
|
|
|
Due to the reimport of the Github repository to bugs.bareos.org, the status of some tickets have been changed. These tickets will be closed again.
Sorry for the noise. |
|
