2

I'm having a problem is sending(creating) an HL7 message using mirth.

I want to read data from my patient table in SQLSERVER 2008 and, using that data, I want to send a message to my destination connector, a file writer. I want my messages to get saved in the file writer's output directory.

So far I'm able to generate the message, but the size of the output file in my destination directory is increasing as the channel's polling time goes on.

Have I done something wrong in the transformer mapping?

UPDATE:

The size of the output file in my destination directory IS increasing. (My .txt file starts from 1 kb and goes to 900kb and so on). This is happening becasue same data is getting generated again and again and multiple times too. for eg. my generated message has one(MSH,PID,PV1,ORM) for one row of data in my Database. The same MSH,PID, PV1 and ORM are getting generated multiple times.

dividius
  • 820
  • 6
  • 16
Rasher
  • 51
  • 2
  • 5

4 Answers4

4

If you are seeing the same data generated in your output directory multiple time, the most likely cause is that you are not doing anything to indicate to your database that a given record has been processed.

For example, if you have 1 record in your database: ["John", "Smith", "12134" ...] on the first poll, you will generate 1 message. If on the second poll you also have a second record ["Fred", "Jones", "98371" ...], you will generate TWO messages - one for John Smith and one for Fred Jones. And so on.

The key is to use the "Run On-Update Statement" of your Database Reader (Source) connector to update the database table you are polling with an indication that a given record has been processed. This ensures that the same record is not processed multiple times.

This requires that your source table have some kind of column to indicate the record has been processed. Mirth will not keep track of this for you - you must do it manually.

dividius
  • 820
  • 6
  • 16
0

You have to have a way of distinguishing what records to pull from the database by filtering on some sort of status flag or possible a time-stamp. Then, you have to use some sort of On-Update statement to mark these same records as processed.

i.e.

Select id, patient, result from results where status_flag='N'

or Select * from results where status_flag = 'N' and created_date >= '9/25/2012'

Then, in either a transformer step or the On-Update section of your Source, you would do something like:

Update results set status_flag = 'Y' where id=$(id)

If you do not do something like this and you have Mirth polling at a certain interval, it will just keep pulling the same records over and over.

Upstart
  • 196
  • 1
  • 2
  • 12
0

You can't have a file reader as a destination, so I assume you mean file writer. You say that "the size of my file in my destination is increasing." Is that a typo? Do you mean NOT increasing?

If it is increasing, then your messages are getting generated and you can view them to start your next round of troubleshooting...

If not, the you should look at the message log in the dashboard to see what is happening on a message-by-message basis - that would be the next place to troubleshoot.

dividius
  • 820
  • 6
  • 16
  • thanks for ur help. but the problem what i'm facing is .....the message keeps on generating( my .txt file starts from 1 kb and goes to 900kb and so on) & the size is increasing continously ,this is happening becasue same data is getting generated again and again and multiple times too. for eg. my generated message has one(MSH,PID,PV1,ORM)for one row of data in my Database , then this 'SAME' MSH,PID, PV1 and ORM are getting generated multiple times. – Rasher Mar 28 '12 at 12:05
-1

You have to change your connector type as Database reader in source.

You have to change your connector type as file writer in the destination.

And you can write your data in the file, For which you have access to write.

while creating HL7 template you have to use the following code in outbound message template

MSH|^~\&|||

Thanks Krishna

kkrgr8
  • 11
  • 4