Hi Viraj/Buddhima,

When doing $subject, I noticed that for a message coming to a proxy with
some payload, at the proxy entry level, only the soap envelope is present,
but not the body of the soap message.
i.e: consider the following message:

*<soapenv:Envelope xmlns:soapenv="http://www.w3.org/2003/05/soap-envelope
<http://www.w3.org/2003/05/soap-envelope>">*
*  <soapenv:Body>*
*    <test>2131312321</test>*
*  </soapenv:Body>*
*</soapenv:Envelope>*

coming to a proxy having only a log mediator and a respond mediator.

so for this proxy we store 3 events each for:

   - the proxy
   - the log mediator
   - the respond mediator

in the two events received for log and respoond mediators, payload look as
follows (which is the expected)

*<soapenv:Envelope xmlns:soapenv="http://www.w3.org/2003/05/soap-envelope
<http://www.w3.org/2003/05/soap-envelope>">*
*  <soapenv:Body>*
*    <test>2131312321</test>*
*  </soapenv:Body>*
*</soapenv:Envelope>*

But for the event received for proxy, the payload looks as follows:

*<soapenv:Envelope xmlns:soapenv="http://www.w3.org/2003/05/soap-envelope
<http://www.w3.org/2003/05/soap-envelope>">*
*    <soapenv:Body/>*
*</soapenv:Envelope>*

As you can see, the body is not there. So a user cannot really search for a
message, at the proxy page (didn't test with a API though). Can you please
have a look?

Thanks,
Supun

On Wed, Mar 30, 2016 at 8:48 AM, Supun Sethunga <[email protected]> wrote:

> Thanks Anjana for the clarifications!
>
> On Tue, Mar 29, 2016 at 11:29 PM, Anjana Fernando <[email protected]> wrote:
>
>> Hi Supun,
>>
>> We should index all the information from the messages, including the data
>> payload as well, where this data will be very valuable in the search
>> operations later. This will not become that much of a big problem, since
>> the indexing happens in the background, and is not in the data publishing
>> critical path. And not to mention, the indexing is also pretty fast. Also,
>> disk space is a different concern, and most probably is not a big problem,
>> and for a very large number of messages throughout a large period of time,
>> if needed, we can of course configure data purging. And also, yeah, when
>> data is purging, it will also delete the corresponding index entries as
>> well, so the raw data in the database, and also the index data as well will
>> be deleted.
>>
>> Cheers,
>> Anjana.
>>
>> On Tue, Mar 29, 2016 at 10:48 PM, Supun Sethunga <[email protected]> wrote:
>>
>>> Hi all,
>>>
>>> As came-up during previous meetings, there was a requirement to do
>>> $subject, when we list down the messages come to a Proxy/API/etc.. in a
>>> given time range. If we are to allow $subject by *payload*, we need to
>>> index the payloads, at esb-analytics (DAS) side. And understandably, since
>>> payloads can be pretty large, indexing payloads for each and every message
>>> would eat-up the disk space (and any possible performance overhead? may be
>>> Gimantha can fill in..).
>>>
>>> Thus, are we going ahead with searching the messages by payloads, or do
>>> we have any alternative search mechanism?
>>>
>>> Also, would like to know, when we are purging the raw data in DAS (say,
>>> per-event-data came few months back), does it also clear any indexed
>>> information as well, which originated from the purged data?
>>>
>>> Regards,
>>> Supun
>>> --
>>> *Supun Sethunga*
>>> Software Engineer
>>> WSO2, Inc.
>>> http://wso2.com/
>>> lean | enterprise | middleware
>>> Mobile : +94 716546324
>>>
>>
>>
>>
>> --
>> *Anjana Fernando*
>> Senior Technical Lead
>> WSO2 Inc. | http://wso2.com
>> lean . enterprise . middleware
>>
>
>
>
> --
> *Supun Sethunga*
> Software Engineer
> WSO2, Inc.
> http://wso2.com/
> lean | enterprise | middleware
> Mobile : +94 716546324
>



-- 
*Supun Sethunga*
Software Engineer
WSO2, Inc.
http://wso2.com/
lean | enterprise | middleware
Mobile : +94 716546324
_______________________________________________
Architecture mailing list
[email protected]
https://mail.wso2.org/cgi-bin/mailman/listinfo/architecture

Reply via email to