I’v been testing LibreNMS for a 6 months now and among other things i’v been monitoring couple of Nokia ISAM. Some of them with thousands of customers and some with only few hundreds.
Lately i discovered that LibreNMS puts a lot of “pressure” on the ISAM and causes sluggish reponse in the CLI. I use default 5 min polling. I have also used config.php to exclude interfaces that are not of interest to me.
Has anybody else experienced this and do you have any suggestion what to do?
Among other things, I’m monitoring 2 geographically separate 7342 ISAM deployments. One has 3 OLT shelves, the other has 7 OLT Shelves. Polling takes forever. I’m still pretty new to Libre. I need to slim it down a lot. We have a similar issue with monitoring/polling our Calix E7-2 shelves, which is similar in functionality.
I don’t have a suggestion for you, but I’m looking for some recommendations myself. Right now, I’m going to just add the shelves as Ping Only device. Eventually I’ll turn on selected port polling and enable the important ports.
Did you notice that ISAM CLI performance is degraded because of this? High CPU usage
I can let you know on that. We had the Nokia ISAM monitored by PRTG for years. It worked, but we’ve recently implemented Libre and I just added the ISAM shelves in (w/full discovery and polling) earlier this week. They kept setting off the +5min polling time alerts. I’ve deleted them and readded as ping only devices for today. I will see if the cli performance and cpu usage goes up after I start adding more sensors in.
I have turned on selective polling for some systems and polling time was reduced by 60% which is great.
Thanks for that tip. I’ve not gotten to test our 7750 performance yet. I will try to soon and reply here with an update.