Periodically today I've encountered this message from the server using BOINC 5.10.20:

Scheduler request failed: HTTP internal server error
Message from server: Project encountered internal error: shared memory
Deferring communication for 1 hr
Reason: project is down


I have been seeing these error messages with a seeming increasing frequency since the issue of newer work units -- particularly the 'mini-workunits' that seem to be getting issued. I'm wondering if there is a server loading or I/O problem -- and if so, could someone shed some light on this someplace?

I am not having problems with other projects -- at least not unannounced problems (Rosetta is down at the moment due to maintenance as noted on their home page, and Predictor has been pretty much out to lunch for several months -- but those are known conditions).

I haven't seen any comment on problems over here though.