Dear OSF1 managers
Last week I posted an inquiry about NFS problems and asked if people knew
of any bugs in NFS2. I was told there was not. It turns out that not only
is the NFS performance sluggish, but general network performance is pretty
bad.
I am running OSF/1 V3.0 on an AlphaStation 200 4/166 with a "Tulip (TU)"
ethernet interface. I have already installed a patch provided by DEC
to correct some known problems in the TU device driver. But the network
performance continues to suffer. In particular, I get many many receive
errors. In particular, in response to a "Netstat -is" command, the system
gives:
tu0 Ethernet counters at Wed Feb 1 11:31:56 1995
65535 seconds since last zeroed
695011760 bytes received
36922439 bytes sent
1825491 data blocks received
520668 data blocks sent
351145733 multicast bytes received
1253100 multicast blocks received
0 multicast bytes sent
0 multicast blocks sent
14746 blocks sent, initially deferred
3676 blocks sent, single collision
3613 blocks sent, multiple collisions
1 send failures, reasons include:
0 collision detect check failure
65535 receive failures, reasons include:
Block check error
Framing Error
Frame too long
0 unrecognized frame destination
0 data overruns
0 system buffer unavailable
0 user buffer unavailable
Does anybody know what might be wrong?
I thought about building NNSTAT and NFSWATCH to monitor my local segment
of the ethernet, guessing that perhaps the problem is not with OSF/1 but
with the network here in Utah. But I cannot find
/usr/examples/packetfilter/pfopen.c which is needed for both.
Can anyone help me with this problem?
thanks
Charlie Jui
Received on Wed Feb 01 1995 - 14:23:47 NZDT