[AGENT++] [Fwd: RE: snmpv3 - question on usm use of engine_id , time window]

[Jochen Katz]

Hi,

> > On scenario TWO below we tried using v3MP::remove_engine_id  which
> > delegates to USM::remove_engine_id
> > and clears the usm_time table as well , so it has the same effect as
 > > the USM::remove_time_information.

yes, but you have to compute the localized keys afterwards.


> > It does not solve the problem , could you suggest
> > another way or workaround
> > to clear the internal state per enginedID so as to make this work.
> > Note: the error via wire shark seems to indicate  that the agent is
> > sending usmStatsNotInTimeWindows report errors.

It would be interesting, what engine boot and engine time values snmp++
sends to the agent and what the values are in the response. Does snmp++
try a second request? If yes, what are the values in the request and in
the response?

 > we are using : SNMP++v3.2.21a , could this have been fixed in a later
 > release ?

As getting the engine boot/time values from the agent is a very basic
operation, this should even work with the oldest snmpv3 enabled snmp++.

3.2.22 has a changelog that says "- Fixed: Check for duplicate entries
in engine id table.". Unfortunately I cannot say what this was about...
If you want to be definitely sure, you have to update to the latest version.

Regards,
   Jochen