Description of problem:
We using the OSP director to deploy the overcloud and we adding our custom features and configurations via heat stack-update. Those changes executed by custom puppet classes and one of those configures the journal to use persistent storage with the following code:
service { 'systemd-journald':
ensure => 'running',
enable => 'true',
}
file { '/etc/systemd/journald.conf.d':
ensure => 'directory',
owner => 'root',
group => 'root',
mode => '0755',
}
file { '/etc/systemd/journald.conf.d/xxxxxx.conf':
ensure => 'present',
require => File['/etc/systemd/journald.conf.d'],
owner => 'root',
group => 'root',
mode => '0644',
content => "[Journal]\nStorage=xxxxxx\n",
notify => Service['systemd-journald'],
}
This puppet file executes successfully and restarts the systemd-journald on every node but somehow the puppet execution stops. Many heat resources left in create_in_progress state and doesn't do anything when in a normal execution those are executed in a few seconds. The stack update fails eventually after a few hours when the keystone token expires. Then those resources state goes to create_failed, but when I check the resource with heat deployment-output show I don't get anything(check the commands.txt).During debugging the problem I noticed that if I removed the notify then the stack update would finished successfully.
Version-Release number of selected component (if applicable):
OSP7
$ grep heat */installed-rpms | awk '{print $1'}
overcloud-controller-0.localdomain/installed-rpms:openstack-heat-api-2015.1.4-1.el7ost.noarch
vercloud-controller-0.localdomain/installed-rpms:openstack-heat-api-cfn-2015.1.4-1.el7ost.noarch
vercloud-controller-0.localdomain/installed-rpms:openstack-heat-api-cloudwatch-2015.1.4-1.el7ost.noarch
overcloud-controller-0.localdomain/installed-rpms:python-heatclient-0.6.0-1.el7ost.noarch
undercloud/installed-rpms:heat-cfntools-1.2.8-2.el7.noarch
undercloud/installed-rpms:ncio-tripleo-heat-templates-1.2-1.el7.noarch
undercloud/installed-rpms:openstack-heat-api-2015.1.4-1.el7ost.noarch
undercloud/installed-rpms:openstack-heat-api-cfn-2015.1.4-1.el7ost.noarch
undercloud/installed-rpms:openstack-heat-api-cloudwatch-2015.1.4-1.el7ost.noarch
undercloud/installed-rpms:openstack-heat-common-2015.1.4-1.el7ost.noarch
undercloud/installed-rpms:openstack-heat-engine-2015.1.4-1.el7ost.noarch
undercloud/installed-rpms:openstack-heat-templates-0-0.8.20150605git.el7ost.noarch
undercloud/installed-rpms:openstack-tripleo-heat-templates-0.8.6-127.el7ost.noarch
undercloud/installed-rpms:python-heatclient-0.6.0-1.el7ost.noarch
How reproducible:
Always
Steps to Reproduce:
1. create a heat stack containing the previous puppet manifest restarting journald (attached the stack I used to reproduce)
2. update the overcloud
3.
Actual results:
The update hangs
Expected results:
update finish
Additional info:
also reproduced with OSP9
Hello Steve. Please note that this bug reports the same issue as bug 1315442, but nominates it for different RHOSP version. I am sorry about confusion created by my comment. Re-opened the bug.
BR, Alex.
Description of problem: We using the OSP director to deploy the overcloud and we adding our custom features and configurations via heat stack-update. Those changes executed by custom puppet classes and one of those configures the journal to use persistent storage with the following code: service { 'systemd-journald': ensure => 'running', enable => 'true', } file { '/etc/systemd/journald.conf.d': ensure => 'directory', owner => 'root', group => 'root', mode => '0755', } file { '/etc/systemd/journald.conf.d/xxxxxx.conf': ensure => 'present', require => File['/etc/systemd/journald.conf.d'], owner => 'root', group => 'root', mode => '0644', content => "[Journal]\nStorage=xxxxxx\n", notify => Service['systemd-journald'], } This puppet file executes successfully and restarts the systemd-journald on every node but somehow the puppet execution stops. Many heat resources left in create_in_progress state and doesn't do anything when in a normal execution those are executed in a few seconds. The stack update fails eventually after a few hours when the keystone token expires. Then those resources state goes to create_failed, but when I check the resource with heat deployment-output show I don't get anything(check the commands.txt).During debugging the problem I noticed that if I removed the notify then the stack update would finished successfully. Version-Release number of selected component (if applicable): OSP7 $ grep heat */installed-rpms | awk '{print $1'} overcloud-controller-0.localdomain/installed-rpms:openstack-heat-api-2015.1.4-1.el7ost.noarch vercloud-controller-0.localdomain/installed-rpms:openstack-heat-api-cfn-2015.1.4-1.el7ost.noarch vercloud-controller-0.localdomain/installed-rpms:openstack-heat-api-cloudwatch-2015.1.4-1.el7ost.noarch overcloud-controller-0.localdomain/installed-rpms:python-heatclient-0.6.0-1.el7ost.noarch undercloud/installed-rpms:heat-cfntools-1.2.8-2.el7.noarch undercloud/installed-rpms:ncio-tripleo-heat-templates-1.2-1.el7.noarch undercloud/installed-rpms:openstack-heat-api-2015.1.4-1.el7ost.noarch undercloud/installed-rpms:openstack-heat-api-cfn-2015.1.4-1.el7ost.noarch undercloud/installed-rpms:openstack-heat-api-cloudwatch-2015.1.4-1.el7ost.noarch undercloud/installed-rpms:openstack-heat-common-2015.1.4-1.el7ost.noarch undercloud/installed-rpms:openstack-heat-engine-2015.1.4-1.el7ost.noarch undercloud/installed-rpms:openstack-heat-templates-0-0.8.20150605git.el7ost.noarch undercloud/installed-rpms:openstack-tripleo-heat-templates-0.8.6-127.el7ost.noarch undercloud/installed-rpms:python-heatclient-0.6.0-1.el7ost.noarch How reproducible: Always Steps to Reproduce: 1. create a heat stack containing the previous puppet manifest restarting journald (attached the stack I used to reproduce) 2. update the overcloud 3. Actual results: The update hangs Expected results: update finish Additional info: also reproduced with OSP9