Would like greater visibility for vNode Pool stress and IFS seize/contention

Worked with IBM to troubleshoot an IFS performance issue that grew slowly but then caused production issues for a couple of weeks.  The problem was ultimately found to be use of an old API SYSTOOLS.httpGetClob() that spawns a Java JVM for every call.  The offending code was making calls from multiple programs and menus and caused the JVM count on the IBM i to go from 58 to 1000+.  This rapid increase in JVM count at peak time of day eventually caused IFS performance issues in the form of an increase in general lock or "seize contention" and exhaustion of "Vnodes".  As we struggled to determine root cause it became clear the current tools such as iDoctor Job Watcher and Help Systems Performance Navigator could not track "exhaustion of VvNodes".  

A Vnode is an object in kernel memory that speaks the UNIX file interface (open, read, write, close, readdir, etc.) apparently. Vnodes can represent files, directories, FIFOs, domain sockets, block devices, character devices.  Prior to this issue I had no idea what a VNode was and how to track when it is "stressed".  We had difficulty pinpointing this issue as iDoctor Job Watcher, IFS dumps and MGTOOLS SYSSNAP were showing different processes at the top of the stack each time the IFS performance slowdown occurred.  Through general observance I figured out the issue was the many JVM's that would ramp up during testing of a new process and the processes at the top of the stack were merely victims during the slowdown as they were all writing to the IFS.

Would like IBM to provide enhancements to current iDoctor tools and perhaps the recent SQL services associated with WRKJVMJOB command could be enhanced to allow tracking of Vnode pool exhaustion so we can be more proactive in resolving issues.