|
scotthernandez, we already evaluated a very similar request, and closed it as Won't Fix: see SERVER-16378. I'll quote my comment from that ticket.
After consulting with some of our support engineers, we have concluded that statistics already provided on numbers of page faults is a good proxy for the "indexCounters", "recordStats", and "workingSet" sections. These sections provide insight into whether your data / indices fit into memory, or whether there may be thrashing due to a large working set. In many cases, the same information should be available via page fault counts. Closing as Won't Fix.
Can this ticket be closed as a duplicate?
|