2 of 2 people found this helpful
So the short answer is YES. There are several things like your plugin can be configured with a port you can do a redirect to a proxy port, you can configure a private port in ARS and use it for each consumer.
But jumping to a port issue for slowness isn’t the first place I’d look. Can you provide any more detail? Consumming a WSDL really isn’t much more than a user doing a query, create or update natively, it shouldn’t be any slower than a user doing the same thing. Have you ran any diagnostics on what the issue is? Are you following best practices? SOAP Web Services Best Practices
Aiming at a need for another port doesn’t sound like a solution unless you ran logs and show all the threads are used and it is queuing work. But if your dB or Application has limited threads or bandwidth that isnt going to solve anything, if you are lacking resources it isn’t going to fix it either. Many times when I hear about this, I’ll do the same thing manually and then using the WSDL and I find that the issue is configuration. Many times the Java sizing, garbage collector, or even classpath just isn’t right. Sometimes architecturally, the customer has ONE integration server and 1000’s of transactions coming in through there and that architecture just doesn’t work! A few other times we found that the WSDL call was a query that was causing a huge verbose tablescan, adding an index solved the issue For one of them and another we created an AO integration and just did away with the need entirely.
have you ran WSDL-analyzer? Have you ran fiddler? Have you ran logs?
Thanks for the detail explanation.
For my architecture not following the best practices of webservices which is need the staging forms, i consumed the webservices (set fields actions) directly in the main forms.
I will try to check the Db and application bandwidth to make sure if is it the root cause.
Which java sizing, garbage collector and classpath configuration you mean i need to check?
Is there any documentation that show us how to set up the one midtier as the webservice integration server because right now we are using the high availability architecture but not spesifically set one of the midtier as the webservice integration server? for my case, it's about 500-700 transaction to run this webservice consumption processes.
Added index already into the table but seems the issue is still persist although indexing make it much better than without indexing.
Not yet, i will try to check using the fiddler today.
1 of 1 people found this helpful
so there is documentation for you to look at, but keep in mind that the version of Java really matters before you just dive in.
As to what to Change I’d have to know where the issue or Slowness is: is it getting to the mid tier, is it passing it from the mid tier to the App server, is it the WSDL event, is it the database Or maybe a combination of more than one of these. More than likely it will be at the ARServer but not sure at this point. The articles above give you insight to each in your environment.