Conference netcad::hub_mgnt

3215.0. "900CP Portswitch Excessive Collisions???" by PH4VAX::PUGH () Mon Jan 29 1996 13:24

DECHUB 900, agent module, 900EF, 900CP, 900EE.
Hubwatch version 4.01

Customer installed and configured a 900CP in his DEChub900 in late Oct. 95.
After moving all his file servers, admin and manufacturing to the port switch
900CP the group (Lan #3) with all the file servers start getting excessive
collisions. All manufacturing gets disconnected, and unable to connect to 900CP
via Hubwatch. This has now happened with three 900CP modules, and the customer
will not put manufacturing back on until we have an answer.

    Per customer, 900CP (decpm-ma) boggs down when network traffic is heavy.
    Will start working when load is removed. Customer states this is the third 
    time this has failed. In fact the first failure was a port failure, hardware
    900CP replaced, 2nd one failed simialiar to this failure and 900CP was 
    replaced again. CSC also requested DECAgent module be replace during 2nd 
    failure. All 900CP modules run for a week to 10 days with full network
    load, then everything slows down and starts to get disconnected. Last 2 
    failures all nodes in group with main file server showed 10K+ send failures
    do to exceesive collisions. Approx. send failure rate of once second on all
    nodes on Lan #3 of 900CP.

1)Any known problems with Hubwatch V4.01 and the 900CP?
2)Any known problems with the 900CP?

Thanks in advance
Charlie Pugh