I was going to mention both of those; -If you want external boxes to run tests from join NLNOG Ring. -Also RIPE Atlas Probes and Anchors can be used.
If you take an external VPS somewhere you may be able to get a BGP feed to it, which you can then receive with Bird/Quagga/ExaBGP from the provider of that VPS and then monitor for your self the BGP changes to routes relating to your network. This maybe something you can do with the NLNOG Ring (we're not part of if yet so I don't knot, but I plan to join ASAP!). A cheeky and shameless self plug, but if you don't already know Smokeping I suggest you check it out (http://oss.oetiker.ch/smokeping/) -> Following on from that, (the cheeky bit) I have been writing something called Smoketrace (very originally named!). It's not on GitHub yet but I hope to get it up there Q1 2014; https://github.com/jwbensley/Smoketrace It looks something like this; http://i.imgur.com/ar38AV8.png It performs a traceroute to a destination every 5 minutes and stores the results. The idea is to monitor for route changes in networks other than your own where you typically can't. This is hard to do accurately, so even though the initial alpha release of the code is 90% finished, I need to work on making it more accurate. It will support multiple kinds of traceroute (ICMP, UDP and TCP). You can be alerted when the route chages to a remote host and receive an email with the different hops in the path. If there is a router failure in a peer network close to our peering router with them, would can catch it, if we have Smokeping set up to a similar point in there network and start seeing loss, we could assume congestion and shut down the peering session, as an example. I will hopefully give a lightening talk at the next LINX meeting about it, and at that point put it up onto GitHub as a beta version ready for bug reports et al. Whilst I'm being shameless, I'm looking for some help with the project, if anyone wants to contact me off list to donate some network pipes :) Cheers, James.
