While working with several clients, we found that the Connection Pool Diagnostic Servlet (CPDiagServlet) was not accurately reporting the number of DB connections within the OTM DB Connection Pools. Here's the details:
We've heard that this will be resolved in OTM v5.5 CU05.
--Chris
While troubleshooting performance issues on this server and monitoring it all
day for several days, I realized that OTM appears to have a logical DB leak
within the PRIMARY_JTS connection pool.
.
Throughout the day, the CPDiagServlet shows the total number of connections
slowly increasing, starting at 100 and ending around 154 - 170 (it was 154 on
the day the attached screenshots were created).
.
Please refer to the following forum thread for additional information:
.
.
Per Development, this may be a minor bug. But since the growth was in the
JTS pool, it may be a bug in the container. We can take a look to see if
we're using the wrong management bean metric.
day for several days, I realized that OTM appears to have a logical DB leak
within the PRIMARY_JTS connection pool.
.
Throughout the day, the CPDiagServlet shows the total number of connections
slowly increasing, starting at 100 and ending around 154 - 170 (it was 154 on
the day the attached screenshots were created).
.
Please refer to the following forum thread for additional information:
.
.
Per Development, this may be a minor bug. But since the growth was in the
JTS pool, it may be a bug in the container. We can take a look to see if
we're using the wrong management bean metric.
--Chris