| Author |
Message
|
| vinod_sri |
 Posted: Thu Apr 17, 2008 7:54 am Post subject: Split the message by specific no.of.records |
 |
|
Apprentice
Joined: 10 Dec 2003
Posts: 42
Location: Atlanta
|
Hello,
I have a task to split a bulk message that consists of 100's of transactions (in TDS format) into small messages with 20 transactions per message. Each transaction has a fixed length of 1200 bytes. This shouldn't change after the split. I am able to split the big message into small messages with 1 transaction per message by using the Mapping Node. Can you please provide me a solution for this? this is sort of urgent.
Thanks
Vinod. |
|
| Back to top |
|
 |
| jefflowrey |
| Posted: Thu Apr 17, 2008 8:34 am Post subject: |
|
|
Grand Poobah
Joined: 16 Oct 2002
Posts: 19981
|
We don't do "solutions".
Use a loop and propagate.
Or a submap.
_________________
I am *not* the model of the modern major general. |
|
| Back to top |
|
|
| vinod_sri |
| Posted: Fri Apr 18, 2008 12:10 pm Post subject: Split the message by specific no.of.records |
|
|
Apprentice
Joined: 10 Dec 2003
Posts: 42
Location: Atlanta
|
Jeff,
Thanks for the reply. Basically I am just mofifying the existing flow. As I mentioned earlier, the big message consists of multiple transactions.
I used the mapping Node to split the large message and created one small message for each transaction. But, it takes one commit for each transaction. Now, I am planning to split the large message on 20 transactions a message basis. But I have no clue how we can acheive this in the Mapping Node. Any suggestions, Please
Thanks
Vinod |
|
| Back to top |
|
|
| jefflowrey |
| Posted: Fri Apr 18, 2008 12:14 pm Post subject: |
|
|
Grand Poobah
Joined: 16 Oct 2002
Posts: 19981
|
Map node has a "for" loop.
_________________
I am *not* the model of the modern major general. |
|
| Back to top |
|
|
| kimbert |
| Posted: Sat Apr 19, 2008 12:47 am Post subject: |
|
|
Jedi Council
Joined: 29 Jul 2003
Posts: 5543
Location: Southampton
|
| Why not use a Compute or JavaCompute node? The Mapping node is best for structured data - you're trying to manipulate a BLOB. |
|
| Back to top |
|
|
|
|
