3 Replies Latest reply on Dec 13, 2010 9:29 AM by robertcbrowning

    Indirect monitoring


      I currently evaluating Orion NPM and APM. I need to be able to perform indirect monitoring. We have remote servers at different locations and each server has 2 networks cards and is in fact on 2 networks. We need to be able to monitore some stations on the other network passing by the server. I know it possible with nagios but is there any way I can do this with NPM ?


        • Re: Indirect monitoring

          Orion typically needs to be able to talk directly to the server that it's monitoring.  There may be a few exceptions to this.

          Without better understanding your environment it's hard to suggest a solution.  Could you try and explain your environment and what you are trying to accomplish in more detail?

            • Re: Indirect monitoring

              Thanks Byrona for your response.

              Our production is done by contract manufacturers around the world. So, on each site, we have a server on our network and also on the manufacturer network (using 2 network cards). A serie of production stations are located on the manufacturer network and reported to the server. Each site's server transfers production datas on our master server. So, the goal here is to monitor not only servers but also the tests stations on the manufacturer's network using the server as entry point.

                • Re: Indirect monitoring


                  Just a couple of quick ideas.

                  1. You could install more NPM/APM primary servers at the remote sites and then use the Enterprise Operations Console to pull all the data together. You could do all the required tests with this option.

                  2. You could add additional engines to your existing primary polling engine and locate them remotely. This would give a similar view to option 1, but you'd need good WAN connections.

                  3. You could use the IPSLA module functions to do some remotely initiated tests, but these are primarily for WAN performance.

                  You have to work out exactly what it is you need to monitor to select the best option. It may be that you could design your testing to be naturally hub/spoke based and without the additional complexity and expense.