|
Home > Archive > BizTalk Server Orchestration > September 2004 > Strange Transactional Behavior Please HELP
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
Strange Transactional Behavior Please HELP
|
|
| Patrick Wellink 2004-09-29, 8:02 pm |
|
I have a simple orchestration that has a dynamic send port
The send port writes the file to disk.
The next shape is an expression shape that will pickup the file
And unzip it to a specified location…..
Now the funny thing……
1. If I execute a io.file.exists(“blabla.zip”) in a expression shape after
the filesend , the file exists ….
2. If I execute a io.file.exists in a component (created in a expression
shape) after the file send, it doesn’t exist
3.
This has probably to do something with transactions…..
But now the funny thing………
If I put a timershape immediately after the filesend the orchestration wil
dehydrate and the file is there……..
I am lost…..
I have the following questions
1. Can anyone explain this behaviour
2. Does dehydration indeed break transactional behavior
3. How long does a timespan have to be before Biztalk decides it wil
dehydrate….
| |
| Patrick Wellink 2004-09-29, 8:02 pm |
| Found the error.....
io.file.exists(“blabla.zip”)
should be
io.file.exists(“blabla.zip”) == true
still strange you have to test for the file in a loop. If you test
immediately after the send (with delivery not. = DELIVERD) you would expect
the file is on disk after the send completes....
But there is a space of about 2 seconds where the send is finished and
the file is not yet there......
"Patrick Wellink" wrote:
>
> I have a simple orchestration that has a dynamic send port
> The send port writes the file to disk.
>
> The next shape is an expression shape that will pickup the file
> And unzip it to a specified location…..
>
> Now the funny thing……
>
> 1. If I execute a io.file.exists(“blabla.zip”) in a expression shape after
> the filesend , the file exists ….
> 2. If I execute a io.file.exists in a component (created in a expression
> shape) after the file send, it doesn’t exist
> 3.
> This has probably to do something with transactions…..
>
> But now the funny thing………
>
> If I put a timershape immediately after the filesend the orchestration wil
> dehydrate and the file is there……..
> I am lost…..
>
> I have the following questions
>
> 1. Can anyone explain this behaviour
> 2. Does dehydration indeed break transactional behavior
> 3. How long does a timespan have to be before Biztalk decides it wil
> dehydrate….
>
>
>
|
|
|
|
|