utl_http efficiency over utl_dbws

One last question regarding utl_dbws:

I just tested the callout sto a webservice over https, comparing utl_http calls with utl_dbws calls. I found that processing times were more than twice as fast with utl_http over utl_dbws.  

Has anyone else experienced this? This could be a deciding factor in what to make the standard for development in our team for callouts.  Are there certain things to consider regarding the environment that may be causing utl_dbws to be so much slower?
Melodi RobertsProgrammer/AnalystAsked:
Who is Participating?
sdstuberConnect With a Mentor Commented:
utl_dbws parses your inputs and constructs a response to send to the service.
Even in when you send a pre-constructred XMLTYPE input, that input is then parsed and reconstructed into a new xml string that is sent.  There are other validations on the inputs too.

All of that work is overhead that you don't incur with utl_http.

utl_dbws.set_logger_level('ALL')  and dbms_java.set_output
  to turn on debugging for your utl_dbws code and you can see logs of some of the work it does
the output will be written to the dbms_output buffer
Melodi RobertsProgrammer/AnalystAuthor Commented:
Thanks so much for the info.  We're probably going to develop a generic package for soap communications using utl_http instead of using utl-dbws.  When processing over 10K records, that extra overhead really adds up.
Melodi RobertsProgrammer/AnalystAuthor Commented:
Just an additional note:  Found a generic soap package using utl_http already written and available at this site: http://www.oracle-base.com/articles/9i/ConsumingWebServices9i.ph
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.