Hi,
I'm using copy xml feature (return as xml) in message mapping to get the xml content into a string. and passing the string through a JAVA UDF to compress it using gzipoutputstream class. This interface is processing around 500 messages per day, and one message size is 7MB. Message processing and data compression works correctly without any issues. But we have noticed PI database is growing very fast after this solution is implemented. only difference in message mapping is, use of a Java UDF to do the compression. all other field mapping including "return as xml" feature is same as before. Message archiving and deletion is activated, background jobs are working without any errors as usual. After this solution is implemented , database is running out space after three days of processing. We're on PI7.3 AEX (Java Only version). Retention period for processed messages is 1 day. Database should be able to hold message load from one day as it worked before. Following is the JAVA UDF, I'm using. Could you please let me know, if this can cause any issues in PI database. Any sugessions or ideas, greatly appreciated.
public String msgcompress(String str, Container container) throws StreamTransformationException{
ByteArrayOutputStream out = new ByteArrayOutputStream();
try
{
//Compress Message
GZIPOutputStream gzip = new GZIPOutputStream(out);
gzip.write(str.getBytes("UTF-8"));
gzip.close();
byte[] theZByteArray = out.toByteArray();
//Convert Compressed binary into BASE64 string
str = DatatypeConverter.printBase64Binary(theZByteArray);
}
catch (IOException uee) {
uee.printStackTrace();
}
return str;
}
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.