I've recently installed an instance of PRPC 7.3.1 on a intranet VM, using SQL Server and Tomcat 8.5. This is my first Tomcat install from scratch, and as the system can be very slow at time with only me logged on, I feel like there are things I'm doing wrong/ways to improve this in the Tomcat conf files (setenv.bat, setDomainEnv.bat, server.xml, web.xml).
In setenv.bat, I edited the catalina and java options based on online advice to be: set CATALINA_OPTS=%CATALINA_OPTS% -Xms4096m -Xmx8192m set JAVA_OPTS=%JAVA_OPTS% -server
In server.xml, I edited the Connector based on online advice to be: <Connector port="80" protocol="org.apache.coyote.http11.Http11Nio2Protocol" connectionTimeout="20000" redirectPort="443" maxThreads="400" minSpareThreads="40" maxKeepAliveRequests="1000" compression="on" compressableMimeType="text/html,text/xml,text/plain"/>
In web.xml, I edited a few servlet params based on online advice to be: development = false trimSpaces = true
For reference, the VM running Tomcat is separate from the VM running the database server. The Tomcat VM is a 64bit Windows system with a single 2.6GHz processor and 12GB of RAM.
Any better ways to do this? I've of course referenced the installation guide on the PDN, which doesn't go into much detail on this infrastructure side of things.
***Edited by Moderator Marissa to update SR Details***
I actually raised SR-C9273 prior to creating this post, and was told to ask the Product Support Community as it was deemed outside the scope of GCS services..
My main concern is that I really don't know much about the varying performances of each connector, nor am I wholly familiar with the available tomcat startup parameters/consequences of using them, like -server which I read can improve performance, or what thread counts people have found work best. So I was hoping I could receive some direction on these fronts from those who have set up environments in the past.
What do you have for the DB connection config in Context.XML? What does it look like? Can you share that? Also what is the version of MS/SQL that you are using. There are some issues with later versions of the SQL Server optimizer that we needed to work around.