Since you do not micromanage the system, my suggestion is to reboot the system which should set any system setting back to normal. If you are running Windows, there have been reports in the past of memory leaks which cause constipation in the system when memory from finished jobs does not get properly freed when the task finishes, thus leading to memory bloat.
If you could send a listing of recent messages that would also be helpful.
Cheers

Since those errors & no replies, rig downloaded tasks on 7th Aug. 145 & 38 @20:45 UTC. Looking at BM the next day, saw what I thought were way too many tasks. Checking via event log saw that prior to those downloads, it had downloaded 134 at 20:43 UTC.

That's a total of 317 tasks & felt pretty certain that a fair number will time out (just my perception). With 29 remaining, 21 waiting, 8 crunching, 4 would have no trouble meeting the deadline with (what looked like) 4 completing shortly past their deadlines. Surprisingly, they completed within 2 minutes before the deadline. I actually saw 13 of the remaining get aborted, not started but the remaining 8 had already started crunching, so:
1: do I let them crunch or abort them?
2: any advice as to how to avoid this happening again?

Have not had any issues on any rig with OPN or MIPS.

Thanks, instead of having 6 + 0.01, I now have it at 1 & 0. Only downloaded 60 or so, good enough for me.