If the connection is not stable between the PathSolutions server and the end device (or a bunch of end devices at the same time), then we will show white connections as being that the connections

In general, it’s best to monitor the local end of a WAN link for utilization and status since the latency to/from the local device will be very low versus the latency to fetch the link status & utilization from the far-end of the link.

If all of the links are being monitored are local links, then there really shouldn’t be any problem with polling them every 5 seconds for status.

If the problem continues, there are two things that can be done:

  1. Investigate why the device is not responding to the SNMP queries in a faster period of time (most devices should respond across a LAN within 100ms).
  2. Change the timeout for the map update to wait a longer time for a response from a slow responding device.

If you want to use option #2, there are some registry entries to change:

  1. Run RegEdit
  2. Navigate to HKEY_LOCAL_MACHINE/Software/NetLatency/SwitchMonitor
  3. Add the key (if it does not already exist): DestWebMapUpdate (DWORD)
  4. The default for this is 5000 decimal (5 seconds). Change this to 1000 (10 seconds).
  5. Add the key (if it does not already exist): PollCurrentTimeout (DWORD)
  6. The default value for this is 1000 decimal (1 second). Change this to 3000 (3 seconds).
  7. Stop and restart the PathSolutions TotalView service

This will have the map wait longer for a response and should help clear up the problem.