> On 14 Jul 2015, at 19:21, Clyde Wildes (cwildes) <cwil...@cisco.com> wrote:
>
> Lada,
>
> I would prefer a specification that declares the "all" intent instead of
> relying on "debug" meaning all messages. A number syslog implementations
> specify a severity of "all". Example: rsyslog -
> http://www.rsyslog.com/doc/v8-stable/configuration/filters.html.
>
> I used the union approach because, in an earlier review, Jan Lindblad
> suggested it as a simple method to add all without compromising the RFC
> definition of Severity. His point was that adding a choice is a complication.
OK, I just expressed my opinion. I also note that RFC 5424 says: "Facility and
Severity values are not normative but often used. They are described in the
following tables for purely informational purposes.” So I wonder whether you
really compromise anything by adding the “all” option.
Lada
>
> Thanks,
>
> Clyde
>
>
>
>
> On 7/14/15, 9:03 AM, "Ladislav Lhotka" <lho...@nic.cz> wrote:
>
>>
>>> On 14 Jul 2015, at 17:25, Clyde Wildes (cwildes) <cwil...@cisco.com> wrote:
>>>
>>> Lada,
>>>
>>> Thanks for your review.
>>>
>>> All was not added to the syslogtypes:severity because that would alter the
>>> definition of severity as specified by RFC 5424 in Table 2 on page 10. I
>>> agree that it will simplify the model to do so.
>>
>> The description of the “severity” leaf says:
>>
>> When severity is specified the default severity comparison is all messages
>> of the specified severity and greater are logged unless all is specified
>> which means all severities are requested.
>>
>> Isn’t then “debug” equivalent to “all”, i.e. is the “all” option really
>> needed?
>>
>>>
>>> Please advise.
>>
>> I would do a similar thing that you did for facilities: choice between
>> “severity” (with the type “syslogtypes:severity”) and an empty leaf
>> "all-severities”.
>>
>> Lada
>>
>>>
>>> Thanks,
>>>
>>> Clyde
>>>
>>>
>>>
>>> On 7/14/15, 6:57 AM, "netmod on behalf of Ladislav Lhotka"
>>> <netmod-boun...@ietf.org on behalf of lho...@nic.cz> wrote:
>>>
>>>> Hi,
>>>>
>>>> the example in sec. 4.3 of draft-ietf-netmod-syslog-model-04 is
>>>> invalid: the value of "severity" should be "critical" rather than
>>>> "syslogtypes:critical". Enums are simple strings, not QNames.
>>>>
>>>> Also, I wonder why the type of "severity" is defined (in multiple
>>>> places) like so:
>>>>
>>>> type union {
>>>> type syslogtypes:severity;
>>>> type enumeration {
>>>> enum all {
>>>> value -1;
>>>> description
>>>> "This enum describes the case where all severities
>>>> are requested.";
>>>> }
>>>> }
>>>>
>>>> I think the "all" enum can be simply added to the "syslogtypes:severity"
>>>> enumeration.
>>>>
>>>> Lada
>>>>
>>>> --
>>>> Ladislav Lhotka, CZ.NIC Labs
>>>> PGP Key ID: E74E8C0C
>>>>
>>>> _______________________________________________
>>>> netmod mailing list
>>>> netmod@ietf.org
>>>> https://www.ietf.org/mailman/listinfo/netmod
>>
>> --
>> Ladislav Lhotka, CZ.NIC Labs
>> PGP Key ID: E74E8C0C
>>
>>
>>
>>
--
Ladislav Lhotka, CZ.NIC Labs
PGP Key ID: E74E8C0C
_______________________________________________
netmod mailing list
netmod@ietf.org
https://www.ietf.org/mailman/listinfo/netmod
