One thing I was kicking around with Mahesh is a compromise on Reshad’s problem 
by adding a “stop” action.  It won’t address the organization of the 
destination, but it would allow for one to express this semantic.  Moreover, 
the actions could be turned into identities (instead of an enum) to allow for 
future extensibility here.

What does the WG think of these options (now that we’re in another LC)?

Joe

From: Kent Watsen <[email protected]>
Date: Friday, January 13, 2023 at 07:58
To: Reshad Rahman <[email protected]>
Cc: [email protected] <[email protected]>, Joe Clarke (jclarke) <[email protected]>
Subject: Re: [netmod] I-D Action: draft-ietf-netmod-syslog-model-28.txt
Hi Reshad,

Thank you for explaining.   I share your assessment that it the model may not 
be implementable in rsyslog.  I also cannot fault the model nor advocate a 
change.  It is unknown to me how pervasive the issue may be, but the model did 
go thru a WGLC before, which would've been the time for the various "network 
OS" vendors to flag issues.  Furthermore, there has been no uproar following 
your message below, in fact, I'm the first to respond.  Based on this, my 
assessment (as both Shepherd and Chair) is to proceed the document now.

Kent



On Nov 2, 2022, at 3:35 PM, Reshad Rahman 
<[email protected]<mailto:[email protected]>> wrote:

Hi Kent,

It's not the text, but the way the YANG model is organized v/s rsyslog 
config+behaviour.

The YANG model is organized with collectors at the top. e.g. for remote 
collectors we have a list of destinations, for each destination a facility-list 
(keyed on facility + severity and ordered-by-user) and for each 
facility+severity tuple we have an action: "block" or "log".

rsyslog<https://www.rsyslog.com/doc/master/configuration/actions.html> config 
is not organized the same way as the YANG model: it first matches on 
facility+severity and then the action is a "collector" (e.g. destination or 
logfile) or "stop". "stop" is not the equivalent of "block": once a "stop" is 
hit, the message is discarded. This means if other destinations were meant to 
receive this message, they won't.

So translating/mapping the YANG model to rsyslog config is problematic when 
"block" is used. As per previous disclaimer, I am no rsyslog expert. If there's 
anyone who's managed to make it work....

And JTBC, I'm not saying the model is wrong since it probably matches how 
many/most network OSes behave.

Regards,
Reshad.


On Monday, October 31, 2022, 08:03:50 PM EDT, Kent Watsen 
<[email protected]<mailto:[email protected]>> wrote:


Reshad,

Which text in the draft are you pointing to?

Thanks,
Kent // as Shepherd



On Oct 17, 2022, at 10:33 AM, Reshad Rahman 
<[email protected]<mailto:[email protected]>> wrote:

Hi,

I believe this model is hard (impossible?) to implement with rsyslog since with 
rsyslog as soon as a message is blocked/discarded, no further processing of 
that message takes place (so other destinations won't get the message either). 
I don't have a solution proposal, just an observation...

Disclaimer: I'm not a syslog expert and I have no idea what implementations out 
there typically do.

Regards,
Reshad.

On Tuesday, October 11, 2022, 01:11:26 PM EDT, Joe Clarke (jclarke) 
<[email protected]<mailto:[email protected]>> 
wrote:



This revision does a few things:


·         Addresses comment from 114 to use 
ct:asymmetric-key-pair-with-cert-grouping instead of 
ct:asymmetric-key-pair-with-certs-grouping

·         Fix Mahesh’s email

·         Replace obsolete RFC references

·         Adjust some line lengths


This passes YANG validation and IDNITS and addresses all known open comments.


We’d like to ask the chairs to conduct another WG LC for this work.


Joe


From: netmod <[email protected]<mailto:[email protected]>> on 
behalf of [email protected]<mailto:[email protected]> 
<[email protected]<mailto:[email protected]>>
Date: Tuesday, October 11, 2022 at 13:04
To: [email protected]<mailto:[email protected]> 
<[email protected]<mailto:[email protected]>>
Cc: [email protected]<mailto:[email protected]> 
<[email protected]<mailto:[email protected]>>
Subject: [netmod] I-D Action: draft-ietf-netmod-syslog-model-28.txt

A New Internet-Draft is available from the on-line Internet-Drafts directories.
This draft is a work item of the Network Modeling WG of the IETF.

        Title           : A YANG Data Model for Syslog Configuration
        Authors         : Joe Clarke
                          Mahesh Jethanandani
                          Clyde Wildes
                          Kiran Koushik
  Filename        : draft-ietf-netmod-syslog-model-28.txt
  Pages           : 41
  Date            : 2022-10-11

Abstract:
   This document defines a YANG data model for the configuration of a
   syslog process.  It is intended this model be used by vendors who
   implement syslog in their systems.


The IETF datatracker status page for this draft is:
https://datatracker.ietf.org/doc/draft-ietf-netmod-syslog-model/

There is also an htmlized version available at:
https://datatracker.ietf.org/doc/html/draft-ietf-netmod-syslog-model-28

A diff from the previous version is available at:
https://www.ietf.org/rfcdiff?url2=draft-ietf-netmod-syslog-model-28


Internet-Drafts are also available by rsync at 
rsync.ietf.org<http://rsync.ietf.org/>::internet-drafts


_______________________________________________
netmod mailing list
[email protected]<mailto:[email protected]>
https://www.ietf.org/mailman/listinfo/netmod
_______________________________________________
netmod mailing list
[email protected]<mailto:[email protected]>
https://www.ietf.org/mailman/listinfo/netmod


_______________________________________________
netmod mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/netmod

Reply via email to