[Last-Call] draft-ietf-nmop-terminology-17 ietf last call Artart review

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

 



Document: draft-ietf-nmop-terminology
Title: Some Key Terms for Network Fault and Problem Management
Reviewer: Tim Bray
Review result: On the Right Track

This is the ART Area last-call review. I may not be the ideal reviewer as I'm
not a networking specialist. I'm going to report lots of things that I didn't
understand and it would be perfectly acceptable for almost any of them to just
say that it'd be obvious to the intended audience of network specialists. But
reporting can't hurt.

I will make one general comment: I'm pretty sure that more examples would be
helpful.

1. Introduction

"Network management comprises a virtuous circle..." I've seen plenty of
networks which were "managed" without any of this stuff. Not effectively or
well. So Perhaps begin the sentence with "Effective Network management
comprises…". Or, given the ending of the previous sentence, maybe just "This
requires a virtuous circle..."

3.1 I didn't understand "categorized into network planes"

3.2 "Characteristic" - this subsection might benefit from an example? Also
applies to other terms defined in 3.2.

Does a Characteristic necessarily have a name or other identifier? Feels sort
of necessary. And wouldn't it have an expected type?
Numeric/boolean/enumerated-value?

3.2 "Event" - suppose I get a 4x packet-rate surge over the course of ten
seconds. Per this definition, this could not constitute an Event? I guess
that's a Change not an Event?  This definition feels shaky because by
definition you can't have a change at a moment in time, there has to be a
delta-T to observe a change. Do you really need both Change and Event?

Ah, upon reading section 4, I get the impression that an Event can be when a
Change is noticed/reported? Maybe say that here?

3.2 "Condition" - "Thus, it is the output of a function". Hmm, the word
"function" has a mathematical meaning (which I think works here) and a software
meaning (which could also work here). Is the definition meant to constrain how
the condition is detected/generated? I think "interpretation" in the 1st
sentence is nice and clear and it's not obvious the second sentence adds value.

3.2 "State" - the border between State and Condition feels fuzzy. Is a State
just a selection of Characteristics/Values?

Also, an example of a Condition is "low available memory" and then the
definition of State includes "that may determine the State of the router, such
as shortage of memory." Problem?

3.2 "Fault" - I didn't understand the explanation of the difference betwen
Fault and Problem. In particular the phrase "from the perspective of managing
the network resources".

3.2 "Problem" - does the existence of a Problem regularly or even necessarily
require human judgment? The language suggests so.

3.2 "Cause" - so a Change can't give rise to a Fault or Problem? Just Event(s)?

3.2 "Symptom" - a Change can't be a Symptom?



-- 
last-call mailing list -- last-call@xxxxxxxx
To unsubscribe send an email to last-call-leave@xxxxxxxx




[Index of Archives]     [IETF Annoucements]     [IETF]     [IP Storage]     [Yosemite News]     [Linux SCTP]     [Linux Newbies]     [Mhonarc]     [Fedora Users]

  Powered by Linux