Announcement

Collapse
No announcement yet.

Nagios / Icinga Monitoring

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Nagios / Icinga Monitoring

    Hi,

    Has anyone out there done any work on nagios plugins to monitor mirth queued messages or errors that they would be willing to share?

    Thanks.

  • #2
    I have a perl script that uses snmp polling to check for errors, number of queued messages & inactivity on a channel (nothing received after x minutes). This relies on the appliance however (uses the appliance mib). Happy to share if it helps.. it executes a bash script for alerting so I guess you could modify this to integrate with nagios.

    Comment


    • #3
      Thanks glenn71 but it probably won't help us here. We don't have the luxury of the appliance which would be easier to monitor via snmp. I have a perl script that runs psql queries to check the counters for queued and errored messages, but it's not very sophisticated and just wondered if there was anything out there before I put in more work.

      Comment


      • #4
        Hi glenn71. I'm looking for a kind of nagios script to monitoring the mirth appliance. Can I ask you to share the script to do some test. Thanks a lot in advance

        Comment


        • #5
          If you want to do active monitoring, polling the Mirth DB is probably the easiest way to get info on channels, alerts, exceptions, etc. The tables CHANNEL_STATISTICS and EVENT will be useful for this information.

          I also devised a way of doing it passively if you don't have access to the Mirth DB using the status and statistics controllers:

          http://www.mirthcorp.com/community/w...us+Controllers

          It took a lot of trial and error to figure out how to make it work just right, but you can get at least the status of the channels channel message statistics with this method. Create a channel that poll these controllers and then use some passive method of sending the check result to Nagios. You could call send_nsca (if you have permission to execute commands), or use the HTTP Sender in the destination connector in conjunction with NSCAweb.

          Comment


          • #6
            Originally posted by Piripera View Post
            Hi glenn71. I'm looking for a kind of nagios script to monitoring the mirth appliance. Can I ask you to share the script to do some test. Thanks a lot in advance
            Hi Piripera, did you finally found a nagios script to monitor Mirth ? I am very interested in this script as I'm looking also for way to monitor Mirth.
            Thanks

            Comment


            • #7
              Originally posted by glenn71 View Post
              I have a perl script that uses snmp polling to check for errors, number of queued messages & inactivity on a channel (nothing received after x minutes). This relies on the appliance however (uses the appliance mib). Happy to share if it helps.. it executes a bash script for alerting so I guess you could modify this to integrate with nagios.
              Hi Glenn, would you be able to share your script to monitor Mirth ?
              thanks a lot
              J

              Comment


              • #8
                I have set up the NSCA service on our Icinga server, and am using JSendNSCA in a polling javascript reader/writer to communicate with it.

                See https://code.google.com/p/jsendnsca/

                Comment


                • #9
                  I have found a Workaround: we Monitor a specific Folder for new files.
                  These files where generated by alert msg with a json string to a channel Reader. this channel Reader got a specific filter on messages and a filewriter Destination with the json string to a log file. nagios is intervalled checking this folder and dropping an alert if a new file is written.
                  after a certain time the log file is moved & renamed by a file Reader to a Archive Folder. within this time, nagios has to check the Folder for new log files and drop the alert.

                  than you get only one message from nagios if the error occures and have the full power to filter and send alertes how it fits best for you + you have an additionl history of errorlogs (which you can prune again).
                  Work for a NPO health care provider in Berlin
                  • Current Version of MC 3.4.1.770 with the onboard MySQL Database on a Windows Server 2012 R2

                  Comment


                  • #10
                    Originally posted by glenn71 View Post
                    I have a perl script that uses snmp polling to check for errors, number of queued messages & inactivity on a channel (nothing received after x minutes). This relies on the appliance however (uses the appliance mib). Happy to share if it helps.. it executes a bash script for alerting so I guess you could modify this to integrate with nagios.
                    Hi Glen - Do you mind sharing the script, please. I am also looking to do something similar.

                    Thanks

                    Comment


                    • #11
                      There is a REST API that didn't exist when this thread was started. I believe it was introduced in mirth 3.4. That's the recommended way to poll the server now. You should be able to find information about using it in the official user guide and by searching these forums.

                      If you are logged into the mirth adminstrator, you can click View Client API in the Other menu to the left to open the self-hosted documentation.

                      Comment

                      Working...
                      X