Re: [OPSAWG] my initial review of draft-schoenw-syslog-msg-mib-00.txt

[Juergen Schoenwaelder <[e-mail]>]

On Fri, May 02, 2008 at 03:41:05PM +0200, Bert Wijnen - IETF wrote:

> In general I think a MIB module like this one makes sense and
> will be usefull.
> 
> Some comments:
> 
> - I see 2 writable objects at the beginning of the MIB module.
>   But it is unclear what the expected persistency behaviour is.

The author just did not know and yes this is where some input would
actually be much appreciated. My preference would certainly be to make
these two scalars be stored in nonvolatile memory. So I plan to add

   The value of this object should be kept in nonvolatile memory.

to both DESCRIPTION clauses.

> - Would it not be easier to use SnmpAdminString instead of
>   DisplayString? Do we never expect any of these fields to
>   contain Internationlized names?

We tried carefully to follow what the SYSLOG specs say. Some message
parts are ASCII while others may contain UTF-8.

> - WHen I see:
>          If the first octets contain the value 'EFBBBF'h, then the rest
>          of the message is a UTF-8 string. Since syslog messages may be
>          truncated, the message may contain invalid UTF-8 encodings at
>          the end."
>   I wonder if it would not be better to truncate in such a way that
>   message allways contains valid UTF-8 encodings.
>   What was/is your motivation to not do that?

The SYSLOG specification allows for arbitrary truncation; a SYSLOG
implementation is allowed to truncate without having to parse the
UTF-8 encoding. And since a truncated message may hit the SNMP engine,
I think it is fair to allow to pass it on without trying to "fix" it.
(And robust applications will have to be able to deal with truncated
UTF-8 strings anyway.)

> - I find these somewhat weird:
>       MIN-ACCESS  accessible-for-notify
>       DESCRIPTION
>           "Read write access is not required."
> 
>   The MAX-ACCESS for this (and other) objects is read-only.
>   So the DESCRIPTION clause seems bogus.
>   I guess you mean that one does not need to support READ
> (GET/HETNEXT/GETBULK)
>   access. You have several of those.

Changed to "Read access is not required."
 
> NITs/TYPOs
> 
> - second para sect 1:
> 
>    This document defines an SNMP MIB modules to represent SYSLOG
> 
>   I guess you want to s/modules/module/

fixed

> - page 9
> 
>           For syslog messages without structured data element parameters
>           that were not truncted by the converter, none of the bits is
>           set."
> 
>   s/truncted/truncated/

fixed

/js

-- 
Juergen Schoenwaelder           Jacobs University Bremen gGmbH
Phone: [phone]         Campus Ring 1, 28759 Bremen, Germany
Fax:   [phone]         <http://www.jacobs-university.de/>
_______________________________________________
OPSAWG mailing list
[e-mail]
https://www.ietf.org/mailman/listinfo/opsawg