Hi, everybody!
I have designed a network and a part of it has a loop consisting of five SF300 switches. There is an NMS PC at another side of the network. I configure router and switches correctly and I can ping any host connected to SF300 switches without any problem. But when I ping five SF300 switches, their behavior to the NMS ping is wired. Randomly some of them don't respond to the ping and NMS gets "request time out" or "destination host unreachable ". And the combination of switches which do not respond to the NMS ping, change once in while. For example, sometimes just switch 1 is out of NMS's reach and after a little while without any change in topology or configurations switch 1 becomes normal and then switch 2 and switch 3 don't respond to NMS ping.
I have tested Both STP and RSTP and nothing has changed. Also, I disable STP on 4 switches and let one of them handle spanning tree in the loop..... the result was the same.
there were three conditions in which switches' behavior get normal:
1- When switch 4 doesn't respond to the NMS ping, for instance, if I connect to the switch 4 CLI directly by a laptop and ping the NMS PC, everything gets back OK and NMS can ping switch 4 again.
2- When there is just one SF300 switch in the loop.
3- When there is no loop.
These three pieces of evidence and the fact that there is no problem to pinging a host (which is connected to an unresponding switch) bring to my mind that maybe SF300 has a defect in its management plane to respond to ping.
So... does anyone have any experience or idea about my problem?
Thank you all, Amir
