[Puppet - Bug #23319] Unnecessary service restart when service is down and we set ensure = running and subscribe the service to some file(s) in puppet-2.7
Issue #23319 has been updated by Alex D.
OK, thanks for the info. I'm not using EPEL, this issue with service
refresh/restart exists in puppet-2.7.23 installed from puppetlabs repo:
pre
# puppet --version
2.7.23
/pre
pre
# rpm -qi puppet
Name: puppet Relocations: (not relocatable)
Version : 2.7.23Vendor: Puppet Labs
Release : 1.el6 Build Date: Wed 14 Aug 2013
09:44:16 PM UTC
Install Date: Wed 04 Dec 2013 11:04:08 AM UTC Build Host: rpm-builder-5
Group : System Environment/Base Source RPM:
puppet-2.7.23-1.el6.src.rpm
Size: 3357077 License: ASL 2.0
Signature : RSA/SHA1, Thu 15 Aug 2013 02:10:55 PM UTC, Key ID 1054b7a24bd6ec30
URL : http://puppetlabs.com
Summary : A network tool for managing many disparate systems
/pre
Bug #23319: Unnecessary service restart when service is down and we set ensure
= running and subscribe the service to some file(s) in puppet-2.7
https://projects.puppetlabs.com/issues/23319#change-100989
* Author: Alex D
* Status: Duplicate
* Priority: High
* Assignee:
* Category:
* Target version:
* Affected Puppet version:
* Keywords:
* Branch:
Steps to reproduce:
1) create test.pp file:
pre
service {
nginx:
ensure = running,
enable = true,
subscribe = File[/tmp/test1];
}
file {
/tmp/test1:
ensure = file,
content = kasjakjksjjks;
}
/pre
2) Install nginx and set error_log level to info
3) Make sure nginx is down and /tmp/test1 file does not exist:
pre
service nginx stop
rm -f /tmp/test1
/pre
4) Apply the test.pp file:
pre
puppet apply test.pp
/pre
Result: puppet brings nginx service up and then restarts it.
Result in stdout:
pre
notice: /Stage[main]//File[/tmp/test1]/ensure: defined content as
'{md5}502675e3d4e9692deb2e3b2e262450c1'
notice: /Stage[main]//Service[nginx]/ensure: ensure changed 'stopped' to
'running'
notice: /Stage[main]//Service[nginx]: Triggered 'refresh' from 1 events
/pre
Result in nginx log:
pre
2013/11/29 16:44:38 [notice] 3900#0: using the epoll event method
2013/11/29 16:44:38 [notice] 3900#0: nginx/1.0.15
2013/11/29 16:44:38 [notice] 3900#0: built by gcc 4.4.7 20120313 (Red Hat
4.4.7-3) (GCC)
2013/11/29 16:44:38 [notice] 3900#0: OS: Linux
2.6.32-358.123.2.openstack.el6.x86_64
2013/11/29 16:44:38 [notice] 3900#0: getrlimit(RLIMIT_NOFILE): 1024:4096
2013/11/29 16:44:38 [notice] 3901#0: start worker processes
2013/11/29 16:44:38 [notice] 3901#0: start worker process 3904
2013/11/29 16:44:38 [notice] 3901#0: signal 15 (SIGTERM) received, exiting
2013/11/29 16:44:38 [notice] 3904#0: exiting
2013/11/29 16:44:38 [notice] 3904#0: exit
2013/11/29 16:44:38 [notice] 3901#0: signal 17 (SIGCHLD) received
2013/11/29 16:44:38 [notice] 3901#0: worker process 3904 exited with code 0
2013/11/29 16:44:38 [notice] 3901#0: exit
2013/11/29 16:44:38 [notice] 3940#0: using the epoll event method
2013/11/29 16:44:38 [notice] 3940#0: nginx/1.0.15
2013/11/29 16:44:38 [notice] 3940#0: built by gcc 4.4.7 20120313 (Red Hat
4.4.7-3) (GCC)
2013/11/29 16:44:38 [notice] 3940#0: OS: Linux
2.6.32-358.123.2.openstack.el6.x86_64
2013/11/29 16:44:38 [notice] 3940#0: getrlimit(RLIMIT_NOFILE): 1024:4096
2013/11/29 16:44:38 [notice] 3941#0: start worker processes
2013/11/29 16:44:38 [notice] 3941#0: start worker process 3943
/pre
Strace:
pre
2912 execve(/bin/env, [env, -i, PATH=/sbin:/usr/sbin:/bin:/usr/bin,
TERM=screen, /etc/init.d/nginx, status], [/* 28 vars */]) = 0
2912 execve(/etc/init.d/nginx, [/etc/init.d/nginx, status], [/* 2 vars
*/]) = 0
2926 execve(/bin/env, [env, -i, PATH=/sbin:/usr/sbin:/bin:/usr/bin,
TERM=screen, /etc/init.d/nginx, start], [/* 28 vars */]) = 0
2926 execve(/etc/init.d/nginx, [/etc/init.d/nginx, start], [/* 2 vars
*/]) = 0
2945 execve(/bin/env, [env, -i, PATH=/sbin:/usr/sbin:/bin:/usr/bin,
TERM=screen, /etc/init.d/nginx, status], [/* 28 vars */]) = 0
2945 execve(/etc/init.d/nginx, [/etc/init.d/nginx, status], [/* 2 vars
*/]) = 0
2954 execve(/bin/env, [env, -i, PATH=/sbin:/usr/sbin:/bin:/usr/bin,
TERM=screen, /etc/init.d/nginx, stop], [/* 28 vars */]) = 0
2954 execve(/etc/init.d/nginx, [/etc/init.d/nginx, stop], [/* 2 vars
*/]) = 0
2966 execve(/bin/env, [env, -i, PATH=/sbin:/usr/sbin:/bin:/usr/bin,
TERM=screen, /etc/init.d/nginx, start], [/* 28 vars */]) = 0
2966 execve(/etc/init.d/nginx, [/etc/init.d/nginx, start], [/* 2 vars
*/]) = 0
/pre
Such behavior may break some services deployment (cluster services that start
to synchronize on the first start and then get stopped/started again).
--
You have received this notification because you have either subscribed to it,
or are involved in it.
To change your notification preferences, please click here:
[Puppet - Bug #23319] Unnecessary service restart when service is down and we set ensure = running and subscribe the service to some file(s) in puppet-2.7
Issue #23319 has been updated by Alex D.
Under puppet-3.3 it works as expected - it brings the server up once without
restarts:
pre
Notice: /Stage[main]//File[/tmp/test1]/ensure: defined content as
'{md5}502675e3d4e9692deb2e3b2e262450c1'
Notice: /Stage[main]//Service[nginx]/ensure: ensure changed 'stopped' to
'running'
Notice: Finished catalog run in 0.44 seconds
/pre
Nginx log:
pre
2013/11/29 16:46:31 [notice] 4605#0: using the epoll event method
2013/11/29 16:46:31 [notice] 4605#0: nginx/1.0.15
2013/11/29 16:46:31 [notice] 4605#0: built by gcc 4.4.7 20120313 (Red Hat
4.4.7-3) (GCC)
2013/11/29 16:46:31 [notice] 4605#0: OS: Linux
2.6.32-358.123.2.openstack.el6.x86_64
2013/11/29 16:46:31 [notice] 4605#0: getrlimit(RLIMIT_NOFILE): 1024:4096
2013/11/29 16:46:31 [notice] 4606#0: start worker processes
2013/11/29 16:46:31 [notice] 4606#0: start worker process 4608
/pre
Bug #23319: Unnecessary service restart when service is down and we set ensure
= running and subscribe the service to some file(s) in puppet-2.7
https://projects.puppetlabs.com/issues/23319#change-100269
* Author: Alex D
* Status: Unreviewed
* Priority: High
* Assignee:
* Category:
* Target version:
* Affected Puppet version:
* Keywords:
* Branch:
Steps to reproduce:
1) create test.pp file:
pre
service {
nginx:
ensure = running,
enable = true,
subscribe = File[/tmp/test1];
}
file {
/tmp/test1:
ensure = file,
content = kasjakjksjjks;
}
/pre
2) Install nginx and set error_log level to info
3) Make sure nginx is down and /tmp/test1 file does not exist:
pre
service nginx stop
rm -f /tmp/test1
/pre
4) Apply the test.pp file:
pre
puppet apply test.pp
/pre
Result: puppet brings nginx service up and then restarts it.
Result in stdout:
pre
notice: /Stage[main]//File[/tmp/test1]/ensure: defined content as
'{md5}502675e3d4e9692deb2e3b2e262450c1'
notice: /Stage[main]//Service[nginx]/ensure: ensure changed 'stopped' to
'running'
notice: /Stage[main]//Service[nginx]: Triggered 'refresh' from 1 events
/pre
Result in nginx log:
pre
2013/11/29 16:44:38 [notice] 3900#0: using the epoll event method
2013/11/29 16:44:38 [notice] 3900#0: nginx/1.0.15
2013/11/29 16:44:38 [notice] 3900#0: built by gcc 4.4.7 20120313 (Red Hat
4.4.7-3) (GCC)
2013/11/29 16:44:38 [notice] 3900#0: OS: Linux
2.6.32-358.123.2.openstack.el6.x86_64
2013/11/29 16:44:38 [notice] 3900#0: getrlimit(RLIMIT_NOFILE): 1024:4096
2013/11/29 16:44:38 [notice] 3901#0: start worker processes
2013/11/29 16:44:38 [notice] 3901#0: start worker process 3904
2013/11/29 16:44:38 [notice] 3901#0: signal 15 (SIGTERM) received, exiting
2013/11/29 16:44:38 [notice] 3904#0: exiting
2013/11/29 16:44:38 [notice] 3904#0: exit
2013/11/29 16:44:38 [notice] 3901#0: signal 17 (SIGCHLD) received
2013/11/29 16:44:38 [notice] 3901#0: worker process 3904 exited with code 0
2013/11/29 16:44:38 [notice] 3901#0: exit
2013/11/29 16:44:38 [notice] 3940#0: using the epoll event method
2013/11/29 16:44:38 [notice] 3940#0: nginx/1.0.15
2013/11/29 16:44:38 [notice] 3940#0: built by gcc 4.4.7 20120313 (Red Hat
4.4.7-3) (GCC)
2013/11/29 16:44:38 [notice] 3940#0: OS: Linux
2.6.32-358.123.2.openstack.el6.x86_64
2013/11/29 16:44:38 [notice] 3940#0: getrlimit(RLIMIT_NOFILE): 1024:4096
2013/11/29 16:44:38 [notice] 3941#0: start worker processes
2013/11/29 16:44:38 [notice] 3941#0: start worker process 3943
/pre
Strace:
pre
2912 execve(/bin/env, [env, -i, PATH=/sbin:/usr/sbin:/bin:/usr/bin,
TERM=screen, /etc/init.d/nginx, status], [/* 28 vars */]) = 0
2912 execve(/etc/init.d/nginx, [/etc/init.d/nginx, status], [/* 2 vars
*/]) = 0
2926 execve(/bin/env, [env, -i, PATH=/sbin:/usr/sbin:/bin:/usr/bin,
TERM=screen, /etc/init.d/nginx, start], [/* 28 vars */]) = 0
2926 execve(/etc/init.d/nginx, [/etc/init.d/nginx, start], [/* 2 vars
*/]) = 0
2945 execve(/bin/env, [env, -i, PATH=/sbin:/usr/sbin:/bin:/usr/bin,
TERM=screen, /etc/init.d/nginx, status], [/* 28 vars */]) = 0
2945 execve(/etc/init.d/nginx, [/etc/init.d/nginx, status], [/* 2 vars
*/]) = 0
2954 execve(/bin/env, [env, -i, PATH=/sbin:/usr/sbin:/bin:/usr/bin,
TERM=screen, /etc/init.d/nginx, stop], [/* 28 vars */]) = 0
2954 execve(/etc/init.d/nginx, [/etc/init.d/nginx, stop], [/* 2 vars
*/]) = 0
2966 execve(/bin/env, [env, -i, PATH=/sbin:/usr/sbin:/bin:/usr/bin,
TERM=screen, /etc/init.d/nginx, start], [/* 28 vars */]) = 0
2966 execve(/etc/init.d/nginx, [/etc/init.d/nginx, start], [/* 2 vars
*/]) = 0
/pre
Such behavior may break some services deployment (cluster services that start
to synchronize on the first start and then get stopped/started again).
--
You have received this notification because you have either subscribed to it,
or are involved in it.
To change your notification preferences, please click here:
