[MRTG] hp lan probes
Greg MATTHEWS
i am currently trying to use HP LanProbe III units to monitor an ageing
network with a view to unpgrading to a faster switched network. i have been
forced to use these units because the hubs appear to be too busy to respond to
snmp requests (an estimated 90%+ traffic). unfortunately it appears that the
hp units will give data for a while (last experiment went well for 15 hours)
and suddenly become blind to snmp requests. has anyone had similar problems?
(error message from cron job shown at end of email)
i have checked for cable/socket problems and tried two different units with
the same results.
perhaps i would get more reliable use if i knew the enterprise mib or at least
the oid for traffic stats. can anyone help...?
on a separate note... i asked this the other day but got no respons - does
anyone know how long the data are averaged over or whether they are
instantaneous snapshots of traffic?
thanks in advance for any help..
cheer
GREG
SNMP Error:
no response received
SNMPv1_Session (remote host: "my.lan.probe" [my.lan.probe].161
community: "public"
request ID: 710247739
PDU bufsize: 8000 bytes
timeout: 2s
retries: 5
backoff: 1)
SNMPGET: Failed to reach target: "1:pub...@my.lan.probe". I tried multiple
times!
---------------------------
Greg Matthews (DSO)
Dept. Computer Science
U.C.L.
London WC1E 6BT
g.mat...@cs.ucl.ac.uk
---------------------------
--
* To unsubscribe from the mrtg mailing list, send a message with the
subject: unsubscribe to mrtg-r...@list.ee.ethz.ch
* The mailing list archive is at http://www.ee.ethz.ch/~slist/mrtg
Knut A. Syed
> this cant be right especially if one is monitoring "dumb" units that are
> unable to process their own stats....but then again... i could be wrong...
MRTG calculates it. ((CounterNow - CounterThen) / ElapsedTime).
~kas
Greg MATTHEWS
> > on a separate note... i asked this the other day but got no respons - does
> > anyone know how long the data are averaged over or whether they are
> > instantaneous snapshots of traffic?
kas wrote...
>
> The average is over the sample-interval (i.e. time between samples),
> which usually is 5 minutes.
>
this cant be right especially if one is monitoring "dumb" units that are
unable to process their own stats....but then again... i could be wrong...
GREG
---------------------------
Greg Matthews (DSO)
Dept. Computer Science
U.C.L.
London WC1E 6BT
g.mat...@cs.ucl.ac.uk
---------------------------
Knut A. Syed
> snmp requests (an estimated 90%+ traffic). unfortunately it appears that the
> hp units will give data for a while (last experiment went well for 15 hours)
> and suddenly become blind to snmp requests. has anyone had similar problems?
> (error message from cron job shown at end of email)
I have been monitoring two HP LanProbes for almost a year without any
problems.
> perhaps i would get more reliable use if i knew the enterprise mib
> or at least the oid for traffic stats. can anyone help...?
I am using the "default" target:
Target[xxx]: 1:community@xxx
The results are available at <URL:http://www.nhh.no/it/mrtg/>. The
uptime on the probes are because of a massive power outage.
> on a separate note... i asked this the other day but got no respons - does
> anyone know how long the data are averaged over or whether they are
> instantaneous snapshots of traffic?
The average is over the sample-interval (i.e. time between samples),
which usually is 5 minutes.
~kas