How do you handle large HL7 messages in SMAT DB?

[Mike Keys]

I believe we are going to start routing some HL7 messages through Cloverleaf that contain encapsulated PDF’s that will be decoded/written out as PDF files for import into another system.

The “issue” is the size of these messages and the affect that this will have on SMAT databases. For example, it takes well over a minute to open an inbound message with over 2 MB of ED data.

Questions:

1. Should we even be saving these inbound messages?

2. Is there a way to make opening these messages quicker? (I’ve toyed around the advanced java settings with no luck).

Topic

[Jim Kosloskey]

Mike,

Are the messages in a file or are they coming in a real time (tcp/ip perhaps) connection?

2 Minutes seems like a long time. We experienced really large radiological results with imbedded (ED) base64 encoded PDF and did not experience anywhere the lag you are seeing.

We did not use SMATDB (file SMAT instead) and did not see a serious issue but then again we had all of our SMAT distributed across multiple SAN environments for better performance as I recall.

Of course we may have had a larger server environment.

If they are coming in a file, perhaps copying the file at the dirparse UPoC of the Fileset and use that instead of SMAT. Or test with file based SMAT to see if that is faster and use that instead of SMATDB for that process. Just some thoughts.

email: jim.kosloskey@jim-kosloskey.com 29+ years Cloverleaf, 59 years IT - old fart.

[Mike Keys]

No, these are real time.

[Author]

Replies