8 Replies Latest reply on Dec 14, 2009 5:24 PM by Donald_Francis

    switch ports

    lchance

      If you have Cisco switches where some have literally hundreds of ports, what is the best methodology in Orion to bring any NEW/ACTIVE ports into monitoring? Is there an automated solution in Orion to bring these ports into Orion? It becomes tedious to manually step through List Resources, etc.

        • Re: switch ports
          qle
          I don't think this is possible since Orion doesn't store the unmanaged ports of each node in the database. This sounds more like a feature request.
          • Re: switch ports
            concentramjones

            I wonder if there's a way to cron List Resources -> (Pause for 5 minutes while my 6513's 600+ interfaces are added) -> Select All Active Interfaces?

              • Re: switch ports
                denny.lecompte

                In 9.5, you will will be able to add all interfaces and mark them so that they show as "unplugged" rater than "down".  Would that help?

                  • Re: switch ports
                    rsprim

                    Denny,

                    Will all those "unplugged" interfaces count for licensing purposes or will their status need to be changed to "active" before they count?

                    Robert

                      • Re: switch ports
                        denny.lecompte

                        Will all those "unplugged" interfaces count for licensing purposes or will their status need to be changed to "active" before they count?

                        Yes.  Those interfaces are fully monitored for stats, but instead of "down", they will show as "unplugged".

                          • Re: switch ports

                            Denny,

                            Does NPM poll nodes that have the Unmanaged attribute set?   The chanllenge we have here is that we want to collect statistics for devices that are up, but on our network less than half the ports are utilized.   We have 15,000 ports and counting, but less than half are actually plugged in.   Sizing the system for 15000 elements is a pain when we nominally should be collecting stats on about about 7000. 

                            Couldn't NPM do something like discover each node onece a day, and add ports that are showing as up, and populate a table so we can track the automatically added ports?     Is there anything like that in an upcoming release?

                             

                            Thanks!

                             

                            Tom

                              • Re: switch ports
                                Donald_Francis

                                We monitor all ports of any switches that are deemed important.  Than being usually any "core" switching.  We do not monitor anything else switchwise.  More ports = more elements = more pollers for one and in general we just do not care about stats on a basic L2 switch.

                                 

                                On the flip side of that we put EVERYTHING into NCM for inventory purposes.

                      • Re: switch ports

                        I don't have an exact answer for what you want, but what we do here is mostly just monitor SNMP traps on the individual interfaces, rather than active polling.  Alerts on the traps are run through a script which queries a table in our database that we populated with Switch,Port#,ServerName assignments.  It then queries the SolarWinds database to see if the server associated with the port that issued the trap is unmonitored or not.

                        This way, we effectively get scheduled downtime on our interfaces, even though that isn't something directly supported by SolarWinds.  And, we don't have to directly fiddle with a poller for the interfaces.  Unfortunately, Solarwinds was already getting the traps by the time I started here, so I wasn't involved in the initial setup for that.  I just work on the scripts + database interaction part of it.