We just upgraded all our servers from OES11sp2 to OES2015sp1 on June 17.
Patched servers as a part of the process. One server, holding a
replica of the edirectory, not the master, has experienced gradually
increasing CPU utilitization until to reports 100% utilization on a 2
cpu virtual ESXi machine.

TOP, top line, far right reports gradually increase numbers of threads
where 1 is the available thread, as the backlog increases, the
performance on the server decreasing, including a simple display of the
directory. This server only services word/excel/pdf files. ndsd does
not go down but eventually the server fails to properly respond or
timely respond. Other servers do not appear to be affected.

Since patched on 6/17, I would think the patch has been applied and we
do not no have ndsd dying. Have a SR at this point but they do not
actually know anything with not ETA for a fix.

Any thoughts/ideas on this matter.