I know this has been addressed in other questions, I don't recall seeing an NLB option being brought up. We have currently a fairly large installation with many probes out at customer sites. Having a single point of failure on our monitoring is unacceptable as our services to customers depend on this.
Would it be possible to utilize a network load balance unit to re-direct probe traffic as it comes in? This way the probe traffic can be sent to the master in which PRTG failed over to. I realize the probes collect the monitoring data for a limited amount of time until the master comes back up, but we need alerting on the fly.
Thanks, Josh H.
Josh, the only way that you can load balance is by using Remote Probes. Like you mentioned, if the Core PRTG goes down, these will hold data up to 500Mb or 500,000 transactions.
You can also set up a Cluster where if the main PRTG Core fails, another will take over but the remote probes will not send data to the failover node.
Remote Probes and Clustering
There are plans to change this but we cannot say how long this may take.
Apr, 2013 - Permalink