Conference netcad::hub_mgnt

2279.0. "NETBEUI and DB90 Performance ??" by PRIVAT::OTTO (...this is the last message You will ever hear...) Mon May 15 1995 12:16

Hello,

I have a question regarding the decbridge 90.

A customer has a AXP2100 and a Vax 4000-90 running VMS V6.1
connected with Thinwire.
When he runs tests between the two systems with "DTSEND" 
he gets a certain amount of thruput between the two systems. 
When he puts a decbridge 90 between the two
systems the thruput is  approx. 1%  less than without bridge.
This is ok and is what the customer expected.

In a second test the customer  replaced the vax 4000-90 
with a PC 486dx33 with a DE200 running msdos V6.2 Pathworks 5.1 client
and used the AXP as Pathworks Server for
the PC. (Pathworks V5.1 is installed on the AXP)

He used the diskperformance tool from norton utilities to measure the 
network drive performance. With the bridge between the AXP and the PC 
he gets 30% less performance than without the bridge.
the PC is configured to use NETBEUI as transport protocol. 


I agree that this is a rather reduced view but in the end there is a
difference of 30% between the test with bridge and the one without 

Any ideas or explanations ???


Thanks 

OTTO

    
    Otto,
    
    	I'd look at the characteristics of the network traffic generated by
        each of the utilities used to do the measurement. Stuff like packet
        size, frames per second, etc.. Not an answer but something to
        consider.
    
    Bob