[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: is there any group interest in my MIB proposal?

--- Juergen Schoenwaelder <schoenw@ibr.cs.tu-bs.de>
> >>>>> B Levin writes:
> Bryan> if there's still group interest in my MIB
> proposal, I am
> Bryan> willing to create and release a sample
> implementation (based on
> Bryan> the net-snmp code base).
> Personally, I generally prefer to fix problems where
> they occur rather than patching things around them.
> But this is just one data point and I think at least
> for comparison purposes, it would be very good to
> have your alternate approach available so that
> people actually can make measurements and provide
> hard facts rather than speculations.

I'm not sure why you think its 'patching around
things' to support a hybrid solution.  its been done
by many vendors with great success to-date.

I admit the hybrid snmp-based snapshot & file-transfer
approach is not a purist solution; but sometimes
getting the job done quickly, efficiently and with
minimal disruption to existing systems can justify
non-traditional approaches.

scp/ftp (etc) is already working and deployed. 
snmpv2/v3 is already working and deployed.  its not a
major effort to extend a subagent by adding another
MIB module at the agent level, and even less effort to
add one at the NMS side.

how long do we want to wait for a purist solution to
be demonstrated, implemented and shipped?

if we care about actually getting bulk data shipped
across wires efficiently AND we want this solution to
be adopted by vendors in our lifetimes, can't we
accept a slightly non-traditional approach TODAY while
letting the R&D effort of the more elegant solution
continue in parallel?

do we care about the disruption that occurs when
EVERYONE must upgrade their NMS to be able to use
these agent new operations?  as a network operator, I
would certainly not relish having to upgrade all my
existing tools simply to solve the problem of
efficient data transfer of management information.

after the benefits of this new protocol version are
demonstrated, perhaps it will be a convincing case to
go thru the expense of a whole retooling effort.  but
that's certainly NOT a short-term thing, speaking from
an operations perspective.  in the meantime, a less
intrusive upgrade via MIB modules sure seems to meet
our needs of solving the problem at hand, today.

Do you Yahoo!?
Yahoo! Tax Center - forms, calculators, tips, more