Steve brought up an im****tant point that I did not address: "Concern is
that
what if the batching interval is up in the middle of processing claims
queued by the orchestration? One end result might include two batch
files,
not one."
My thoughts are such:
Yes, having the batching at 1 minute is kind of small, but really, that is
what testing is all about. If you find that your queuing orchestration
takes
too long to consume all of the messages queued up, then you do 1 of 2
things:
Increase your batching time from 1 to 2 minutes etc
Get a faster box, create separate host instances for that orchestration on
its own, and all other processes use a different host instance.
"Eric Stott" <ericstott@[EMAIL PROTECTED]
> wrote in message
news:O72QVYB0HHA.3916@[EMAIL PROTECTED]
>I was hoping someone would have a more elegant solution, and Steve is
100%
>correct about the settings.
> As I was thinking about your problem I thought of a way to accomplish
what
> you are wanting to do...
>
> Why can't you set the batching time at 1 minute and then for the
> particular process that does not need batching, messages go through to
the
> HIPAA accelerator and files are created with little batching (if
multiple
> messages come in within that 1 minute it will batch them together).
> However for your batching requirement, here is the trick - you create a
> singleton orchestration that picks up the first message and then has a
> timer that waits for 50 minutes (or whatever time it has), the the
message
> box is queueing up all of the other messages and then when the 50
minutes
> are up the timer ends, it consumes the messages queued up and sends them
> all thru to the HIPAA accelerator within that 1 minute - and the HIPAA
> accelerator will batch them together like you need.
>
> Eric
> http://blog.biztalk-info.com
>
>
> "steve" <stevesj@[EMAIL PROTECTED]
> wrote in message
> news:eXoADc$zHHA.3564@[EMAIL PROTECTED]
>>
>> The global batching interval (minutes) is the only control one has over
>> outbound batching. There are no other triggers, settings, apis, etc.
>> The batching interval value is stored in esp.ini. The HIPAA service
>> queries this file every 30 seconds.
>>
>>
>>
>> <ndillon45@[EMAIL PROTECTED]
> wrote in message
>> news:1184953419.429449.97080@[EMAIL PROTECTED]
>>> Our requirement is to create a single 834 eligibility file with
>>> multiple ST/SE segments grouped by a particular field within our
>>> query, i.e. "account_id". This solution works just fine:
>>>
>>> 1. Use Multiple Schema
>>> 2. Use XPath to loop thru each group of "account_id"s
>>> 2a. Create message to send to HIPAA ****t for each account_id group
>>> 3. Turn batching on, set batching interval to 50 minutes.
>>>
>>> Since all of these accounts can be generated in under 50 minutes, this
>>> works fine. However, we have other processes that use the HIPAA
>>> Accelerator and the 50 minute batch interval is slowing these
>>> processes down since they would normally only take a few minutes to
>>> generate.
>>>
>>> So, my question: Is there any way to force the HIPAA accelerator to
>>> batch based off a variable or some other trigger? This would allow us
>>> to batch whenever we determine all messages that need to be in the
>>> batch have been sent, thus reducing the time from the last message
>>> sent to the next batch interval.
>>>
>>> Any ideas?
>>>
>>> Thanks!
>>>
>
>
|
