Dear all,
During the OPSAWG/OPSAREA meeting, I will be addressing one issue that
is now on the top of my mind: OPS guidance to the non-OPS community.
Problem 1:
How do we share our OPS knowledge? How do we advice which OPS tools
(NETCONF, SNMP, AAA, ping, syslog, IPFIX, etc...) the community should
be using?
The point is that we have multiple management frameworks, as opposed to
only one in the past: SNMP
You know, it could be as simple as: if you plan on doing configuration,
investigate NETCONF.
But it might more complex:
1. do you have an existing MIB module?
and you don't plan on doing configuration?
then update the existing MIB module
alternatively, if you're thinking about configuration, then
investigate NETCONF.
and for monitoring, if you start from scratch, use NETCONF that can
also report operational status
however, if you have an existing MIB module for monitoring, and you
plan on using NETCONF for monitoring ...
2. I need to do reporting, can you please give the pros and cons of the
syslog versus syslog SDE versus SNMP versus IPFIX versus ...
On regular basis, I have to provide these advice: it's not always easy,
because the answer depends on many factors. However, when asking a few
questions, we're able to limit the options to just one or two choices.
How can we share this knowledge?
This is the perfect job for an OPS advisor, but one OPS advisor in all
WGs doesn't scale.
Problem 2:
What are the OPS recommendations for future developments?
It might appear as the same problem as 1, but the audience is different.
Here, we speak about new charters for example.
Basically, the recommendations have changed over the years.
For example, don't write MIB modules in your charter to please the OPS
ADs, it's not about MIB module only any longer
Provide the use cases and operational requirements, and from there, we
will help selecting the right protocol and data models
There are pieces of information all over the place:
- RFC 5706, Guidelines for Considering Operations and Management of New
Protocols and Protocol Extensions
Appendix A askes interesting questions, helping with the
operational requirements
- RFC 6632, An Overview of the IETF Network Management Standards
It's complete, but way too long. The appendices ask the right question
- There was an attempt to solve the same problem in the past:
draft-harrington-operations-and-management-01
- RFC 1052, IAB Recommendations for the Development of Internet Network
Management Standards, is 25 years old :-)
What are the constraints?
For problem 1:
must be simple and not too long (not a 50 pages document)
can change on regular basis
Solution: maybe a WIKI or a flow chart?
For problem 2:
These are recommendations for future developments: it doesn't
matter if it's not perfect day one.
Solution: maybe update RFC 1052bis
Questions:
- Do you agree on the two problem statements?
- How to solve them? Any brainstorming ideas are welcome.
Regards, Benoit
_______________________________________________
OPSAWG mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/opsawg
