Description of problem: -------------------------- It is planned to make samba as optional package while installing RHSS using ISO and moving samba out of the rhs repo to a different repo. In that case, Nagios server should not monitor SMB if user chooses not to install samba. Auto-discovery could check for `rpm -q samba' to decide whether SMB service should be created or not. Version-Release number of selected component (if applicable): -------------------------------------------------------------- nagios-server-addons-0.1.11-1.el6rhs.noarch How reproducible: N/A Steps to Reproduce: N/A Actual results: N/A Expected results: ------------------ SMB service should not be created if samba is not installed. User should be expected to run auto-discovery after installing samba at a later point in time. Additional info:
If this happens in 3.1 release of RHS, we need to take care of it as well, otherwise it has to happen, when SMB installation is made optional.
Plugin will be changed to look at chkconfig status of SMB. If turned off, plugin will report SMB service status as "Not Configured".
(In reply to Shruti Sampat from comment #0) > Description of problem: > -------------------------- > > It is planned to make samba as optional package while installing RHSS using > ISO and moving samba out of the rhs repo to a different repo. In that case, > Nagios server should not monitor SMB if user chooses not to install samba. > Auto-discovery could check for `rpm -q samba' to decide whether SMB service > should be created or not. > > Version-Release number of selected component (if applicable): > -------------------------------------------------------------- > nagios-server-addons-0.1.11-1.el6rhs.noarch > > How reproducible: > N/A > > Steps to Reproduce: > N/A > > Actual results: > N/A > > Expected results: > ------------------ > SMB service should not be created if samba is not installed. User should be > expected to run auto-discovery after installing samba at a later point in > time. > > > Additional info: 1) Currently it gives critical error message when some of the volumes configured to use smb but the smb is not configured or not running. 2) Returns "OK" when no volumes using smb and the smb is not configured or running. If we disable the service completely if the supporting packages (samba) are not installed, it will not be possible to inform the user when the above mentioned cases occurred. So we will modify the plugin to look at chkconfig status of SMB. If turned off, plugin will report SMB service status as "Not Configured".
And by-default its enabled for the every gluster volume we create.
Patch sent to upstream for review: http://review.gluster.org/10423
If no volume uses smb and smb is not configured plugin will report "SMB is not configured"
When CTDB is configured user will be doing chkconfig smb off, in this case will nagios report that SMB is not configured? can some please explain?
When CTDB is configured user will be doing chkconfig smb off, in this case will nagios report that SMB is not configured. Is this expected?
Moving this bug back because the fix provided is not valid in all the cases. When CTDB is configured on the systems chkconfig of smb will be off to prevent smb coming up during reboot. In this case if CTDB is configured on the system then smb service will be shown as smb not configured where as SMB is configured.
Triage decided to doc-text it and take it up in 3.1.z
The plugin returns the status as follows: 1) if smb is running it will immediately returns the status as 'OK Process smb is running 2) if smb is not running it will check for check for chkconfig smb status and returns the status if its not configured (as 'OK: Not configured') 3) if smbd is not running and chkconfig smb is not true then it will check for any volumes using smb. If any volume uses smb it will return critical error message as ""CRITICAL: Process smb is not running" 4) if smbd not running, check config smb is not true and no volumes exists or no volume uses smb it will return the message as "OK: No gluster volume uses smb" So the smb not configured message will occur any time when there is no smb service running and when chkconfig smb is false. So it looks like a minior issue all about the message. As per command3, its consider that using chkconfig tool for smb is also a part of smb configuration. However as per comment10, the plugin need not necessarily looking for chkconfig status of smb; because there may be some situation where chkconfig smb will be off and the plugin can just bypass the above step2. However the question here is, if chkconfig is not going to be the part of smb configuration then how are we going to start the service automatically on reboots. Do we have a separate ui to restart this service along with or without ctdb configuration if required or the admin has to do it manually.
Thank you for the bug report. However, closing this as the bug is filed against gluster nagios monitoring for which no further new development is being undertaken.
The needinfo request[s] on this closed bug have been removed as they have been unresolved for 1000 days