Bug 785863

Summary:	Verify cumin-database after changes to use systemctl where supported
Product:	Red Hat Enterprise MRG	Reporter:	Trevor McKay <tmckay>
Component:	cumin	Assignee:	Trevor McKay <tmckay>
Status:	CLOSED CURRENTRELEASE	QA Contact:	Peter Belanyi <pbelanyi>
Severity:	low	Docs Contact:
Priority:	low
Version:	Development	CC:	athomas, croberts, esammons, matt, mkudlej, sgraf, tmckay
Target Milestone:	2.3
Target Release:	---
Hardware:	Unspecified
OS:	Linux
Whiteboard:
Fixed In Version:	cumin-0.1.5251-1	Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:	785840	Environment:
Last Closed:		Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Bug Depends On:	785825, 785837, 785840
Bug Blocks:

Description Trevor McKay 2012-01-30 19:34:55 UTC

Note, this will be a transparent change on RHEL.  The cumin-database script should continue to use /sbin/service, though the code surrounding the calls will change.


+++ This bug was initially created as a clone of Bug #785840 +++

+++ This bug was initially created as a clone of Bug #785837 +++

+++ This bug was initially created as a clone of Bug #785825 +++

Description of problem:

The cumin-databse script has legacy calls to /sbin/service for controlling postgresql.  This causes extra "redirecting" messages to print on the screen when script operations are run.  Eliminate these by calling systemctl.

Version-Release number of selected component (if applicable):

cumin-0.1.5137-5

How reproducible:

100%


Steps to Reproduce:
1.  yum install cumin
2.  cumin-database install
3.  cumin-database stop
4.  cumin-database start
  
Actual results:

At various points in the output from the above commands, messages like this will be printed:

Redirecting to /bin/systemctl  start postgresql.service

Expected results:

These messages should not be seen.

Additional info:

This functionality will be added through conditionals in a single script that supports F15+ and RHEL 5+.  As such, it is not just a simple cut and paste operation since platforms without systemd still need to work.

Comment 1 Trevor McKay 2012-01-31 15:44:22 UTC

If QE already has regression tests that cover cumin-database functionality, perhaps this BZ can be dropped for RHEL.

Comment 2 Trevor McKay 2012-01-31 19:45:53 UTC

Fixed in revision 5198.

Test sequence that should exercise all execution paths for affected code.  Note that users typically will likely only be doing cumin-database install, start, and stop from the below set.

On a clean system

$ cumin-database install

This should succeed.  
RHEL 5 should use /sbin/service postgresql start for initialization
RHEL 6 should use /sbin/service postgresql initdb for initialization

$ cumin-database stop
$ cumin-database start 

The following steps can be done to test an execution path through the "configure" command that will prompt the user and perform a restart.  The "create" will leave everything in order.  Note, not strictly necessary since users are unlikely to use these commands but to be complete...

$ cumin-database annihilate  (answer "really" at the "Really?" prompt)
$ cumin-database initialize
$ cumin-database start
$ cumin-database configure   (answer "yes" to restart)
$ cumin-database create

Comment 5 Eric Sammons 2012-10-03 13:27:10 UTC

This BZ can not be flagged for Grid-2.3 as it appears to be related to systemctl which is a futurefeature, currently available in RHEL 7 and Fedora.

Comment 7 Peter Belanyi 2013-01-16 09:05:06 UTC

Verified on RHEL5 and RHEL6, both i386 and x86_64, with cumin-0.1.5648-1