A device is down, but its downstream devices are up:
It could be that the down device was pinged at a different time to those downstream devices that are up. The down device was up when its downstream devices were pinged, but was down when it was then pinged.
traceroute and ping take a finite time to run, and on occasions can return counter-intuitive results.
Intermittent and misleading results:
traceroute and ping send ICMP packets. When a router is overloaded, it may throw away these low priority ICMP packets, which will impact the recorded availability of the router and its downstream devices Where the device is intermittently available, then the impact would also be shown through increased latency values for the device.
Examine a daily availability report for that area of the network. When the suspect router and its downstream devices have a lower availability than surrounding devices, you should investigate the router's performance.
Routers and redundant links:
When two routers are connected to form a redundant link, the type of event raised when one of these routers goes down depends on the path specified in the router table. For example, if Router 3 goes down: