PTP58600 Ping Testing

I have a PTP link that is giving me the following problem.  I setup the slave side, lets say .4 and ping to the unit and get a great response of 1ms.  This will stay constant for 10 min.  I then put in the distant end, .3, so the link is not completed.  My ping jumps to 1000ms and I constantly get bumped out of .4 because it is timing out.  

I have totally blown out the system to put all fresh info into the system.  It does not matter what I do via IP config, cable config the system will run great until I make the link.

Anything would be appreciated.


Just to clarify, after the link is established I am still pinging the .4 side.  I left the ping test going the whole time and right after link established the problems shows up.


If I understand correctly, the situation is like this: You have a PC connected by a wired network to the slave end of a PTP 600 link. When the slave is powered up and the link is down, the ping response time from the PTP 600 ODU is typically 1 ms, as expected. When the link establishes, the ping response time from the slave (still connected through a wired network) increases to about 1000 ms, and remains at this level.

I tried to set up the same experiment. I observed that the response time from the slave was initially consistent at 1 ms. When the master and slave were bringing up the link, the response time from the slave increased to about 50 m for a few seconds. After the link was established, the slave ping time returned to 1 ms, and the master ping time was about 3 ms. None of this was a surprise.

In your network it would be interesting to know if the network was capable of passing significant traffic over the PTP 600 link as soon as the link established. If this was the case, your longer ping time could be because the ping packets were traversing a loaded network. If the pings were sent with low priority this might account for the additional delay.

Another possibility is that the network contained a loop, for example the master unit might have been connected via the wireless link and also via the Ethernet port. In a network with a loop we rely on STP to block a port to create a simple tree. When this occurs it's difficult to predict which route the packets will take.

One more possibility: In a looped network without STP the broadcast traffic may loop indefinitely creating a so-called broadcast storm.

Could any of these possibilities explain your long ping response times?