On Tue, Jul 16, 2013 at 9:09 AM, Rainer Gerhards
<[email protected]>wrote:

> On Tue, Jul 16, 2013 at 9:06 AM, Erik Steffl <[email protected]> wrote:
>
>> On 07/15/2013 11:52 PM, Rainer Gerhards wrote:
>>
>>> On Tue, Jul 16, 2013 at 8:47 AM, Rainer Gerhards
>>> <[email protected]>**wrote:
>>>
>>>
>>>>
>>>> yup, but in a different flavor: I think everyone can pretend things if
>>>> he
>>>> is able to reconfigure rsyslog. I think the point here is to provide the
>>>> necessary facility to rightful users. And if e.g. mmjsonparse always
>>>> parses
>>>> into the root, you have essentially lost, you can't properly work with
>>>> subtrees. I remember we discussed adding a capability to write to
>>>> different
>>>> subtrees, not sure though if we did implement it.
>>>>
>>>>
>>> I just checked: unfortunately we did not yet do that.
>>>
>>
>>   that would solve my problem I think, if the parser would put data into
>> a user specified subtree instead of root of json.
>>
>>   We are using the dev version (latest packages from adiscon ubuntu
>> repository), just about to release it to production (not a huge one, few
>> servers, maybe a million messages per day). Would definitely at least test
>> the version with parsing into a subtree.
>>
>>   curious: why all the discussions only talk about one json tree. Why not
>> have several named json trees?
>
>
> For the same reason you only have one file system tree: the tree is a
> hierarchical namespace, so why add another namespace on top of it?
>
>
>> Then whenever referring to json one would specify which tree to use.
>
>
> You do - it's the top level
>
> !usr!xxx has all usr variables
> !msg!xxx has all msg variables
> !yyy!xxx has all yyy variables
>
> see the different trees?
>
>
>> Technically it's no different than having one tree
>
>
I forgot: it's a *big* difference: you need an additonal resolve to handle
the various top level trees - with all support infrastructure like hash
tables to keep their names etc. All solved without cost and complexity when
a single name space is used.

Rainer

> but mentally it might be easier to manage (and to keep things separated -
>> incoming parsed message, outgoing message, user specified data etc.)
>>
>>
> again: think filesystem (or dns name or... - it's just the regular method
> of choice).
>
> I'll see if I can quickly add that capability to mmjsonparse. Can you
> confirm this is the parser you need?
>
> Rainer
>
>>   thanks!
>>
>>         erik
>>
>> ______________________________**_________________
>> rsyslog mailing list
>> http://lists.adiscon.net/**mailman/listinfo/rsyslog<http://lists.adiscon.net/mailman/listinfo/rsyslog>
>> http://www.rsyslog.com/**professional-services/<http://www.rsyslog.com/professional-services/>
>> What's up with rsyslog? Follow https://twitter.com/rgerhards
>> NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad
>> of sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you
>> DON'T LIKE THAT.
>>
>
>
_______________________________________________
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com/professional-services/
What's up with rsyslog? Follow https://twitter.com/rgerhards
NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad of 
sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you DON'T LIKE 
THAT.

Reply via email to