I've been seeing a lot of packet loss at intermediate hops. Is it possible to set up an alert on each intermediate hop in a route with PingPlotter?
There are very rarely any situations that would call for an alert to be set up on a intermediate hop.
As we've discussed in previous knowledge base articles - the only hop that matters is the final destination.
The intermediate hops can, however, provide a great deal of information that's valuable to your troubleshooting efforts. If you experience any problems at your final destination, you can go back up through the intermediate hops, and see where an issue originally started.
In some cases - these intermediate hops will relay information to the final destination just fine, but may not return information that is specifically targeted for them (in cases where routers may specifically block or down-prioritize ICMP echo requests). Having an alert on each hop in a route would notify you of these situations as if they were a problem. Your goal should always be to make sure your final destination is working well - so you wouldn't want to be alerted on something that isn't actually causing issues at your target.
Take the following screenshot from a PingPlotter trace as an example:
Here, our target has an average 38ms round trip latency, and no packet loss. See the 100% packet loss at hop #2, and the packet loss at hop #4? While both of those hops seem to be relaying information just fine (because our final destination isn't exhibiting any signs of having issues), they're not responding well to information that is targeted at them (ICMP echo requests). Having alerts tied to these hops could send us notifications that may be completely unnecessary - as our target destination is preforming very well.
In this graph - we can clearly see that there's some packet loss occurring starting at hop #8. We don't, however, need an alert on hop #8 to notify us of this. If an alert is set up on the final destination - we'll be notified of the packet loss at hop #8 (and #9 and #10) - because the issues are carrying through to our target.
A good way to think of this is that having an alert set up on only the final destination will watch the entire route for you. Having alerts tied to each hop can create scenarios where PingPlotter would have the opportunity to "cry wolf" - and notify you of situations that may not be indicative of real problems. An alert on only the final target will notify you of anything that actually creates problems at the final destination (and then you can go back through the previous hops to see where the issue began).
This isn't necessarily to say that there aren't situations where an intermediate hop would need to be monitored. These situations, however, are not typical. If you should find yourself needing to monitor a specific hop in a route - an alert *can* be tied to any hop by right clicking and selecting "Watch this host (Alerts)." You can see more information on how to configure alerts in the Creating / Configuring Alerts portion of our product manual.
Article Number: 100 | Last Updated: June 23, 2015
This article has been viewed 1509 times since May 19, 2014
Filed Under: Alerts