T.R | Title | User | Personal Name | Date | Lines |
---|
2946.1 | | NETCAD::DRAGON | | Mon Nov 06 1995 09:34 | 16 |
|
Sam,
The NA means that the feature is not applicable or not supported.
The COPIED counter relates to the fddimibMACCopiedCts MIB object
and the TRANSMITTED relates to the MIB object fddimibMACTransmitCts
(both from RFC 1512). These MIBs are not supported on GIGAswitch/FDDI.
The DIAL for utilization is not supported by HUBwatch at this time.
As for the interface counters being 0, if you could do an SNMP
Trace and see if the GIGAswitch/FDDI is returning 0s, this would
help isolate this problem.
Hope this helps,
Bob
|
2946.2 | What MIB variables show I trace? | JULIET::MOKBEL_SA | | Wed Nov 08 1995 14:14 | 11 |
| Bob,
From DECMCC I can look at the INTERFACE counters and see traffic on
all the ports, none of them returns 0.
What mib variables did you want me to look at fro the bridge counters
FRAMES IN and FRAMES OUT per second?
Thanks,
sam
|
2946.3 | | NETCAD::DRAGON | | Wed Nov 08 1995 15:22 | 18 |
|
Sam,
The MIB variables concerned are:
dot1dTpPortInFrames and dot1dTpPortOutFrames
What HUBwatch does is to take a sample, gather some other MIB info
and then take another sample. Based on a timestamp it calculates
a per second value. If no frames entered or left the port between
samples then the value(s) will be 0, even though the MIB variable(s)
are not 0.
The time bewteen samples may vary depending on network/system load,
but should be within a few seconds.
Regards,
Bob
|
2946.4 | I checked the MIB variables | JULIET::MOKBEL_SA | | Wed Nov 08 1995 18:02 | 16 |
|
Bob,
I looked at the mib variables dot1dTpPortInFrames and
dot1dTpPortOutFrames. They do change on all the port and there is
traffic in and out. Yet HUBWATCH can't report it. on a couple of ports
where I have VAXs connected I get some numbers for INFRAMES and
OUTFRAMES but the numbers are very small and they do not reflect the
actual frames comming in or going out, also, the FILTERED field is 0
for those ports. The only port where I have correct information is the
one that connects to a DECCON 500 (this is a B port, all other ports
are M ports!)
Thanks,
sam
|
2946.5 | | NETCAD::DRAGON | | Thu Nov 09 1995 08:14 | 12 |
|
Sam,
Thanks for the info.. One possibility that I can think of is that
when HUBwatch gets the mibs that it hits a lull in the traffic on
that port. Since it takes a sample, waits a short while, then takes
another sample, there is the possibility that little or no traffic
was seen during the interval. What you might want to try is to
create some constant traffic on that port and then look at what
HUBwatch shows.
Bob
|
2946.6 | makes sense | JULIET::MOKBEL_SA | | Wed Nov 15 1995 18:38 | 11 |
| Bob,
I looked at the bridge port statistics for IN and OUT frames again and
I do have traffic, and the filterd counter is 0 on ports where I have
VAXs only makes sense now. I am not supposed to see any traffic going
into a port where there is a single station unless that traffic is
destined to that station.
Thanks,
sam
|