My clientcontacts GridRepublic and Einstein@Home every hour on the hour, as shown below.

13-Sep-2009 05:23:09 [---] Contacting account manager at http://www.gridrepublic.org/
13-Sep-2009 05:23:12 [---] Account manager contact succeeded
13-Sep-2009 05:23:17 [Einstein@Home] Sending scheduler request: Requested by account manager. Requesting 0 seconds of work, reporting 0 completed tasks
13-Sep-2009 05:23:23 [Einstein@Home] Scheduler request succeeded: got 0 new tasks

I understand (from reading Discussion topics) the rational for the client to contact GridRepublic periodically--to keep updated with changes made on the GridRepublic website. There doesn't appear to be any reason for the client to periodically contact Einstein@Home, however. The frequent nonproductive contacts with E@H just clutter up the Messages tab, making it more difficult to locate relevant information.

Is there a .xml file I could edit to stop these contacts with E@H, such as client_state.xml, sched_request_einstein.phys.uwm.edu.xml or perhaps some other file? On the other hand, there may be nothing users can do, since the scheduler request is made by the account manager. So it's an interesting problem.

If you would kindly tell me if there isa fileI couldedit anda valueI could...

Matt, I have the same annoying problem with Einstein@home as Joachim has with Rosetta. As for your suggestions concerning the reasons this happens is concerned:

"having said that, i'm not sure why each contact with GR initiates a scheduler request with the project. this may be a peculiar result which happens only when there is no work at the desktop for a listed project. and it may simply be that rosetta is out of work for machines of your description.

the answers are no and no. It would be nice and much appreciated if you would eliminate this unnecesary idiosyncracy from your software.

Steve

Thanks for the reply, Spark. When I conduct benchmarks I make sure thereare no other controllable activities on my computer. It could be that the algorithms are different, but since benchmark algorithms are synthetic I may not see any difference in real world computation times. Time will tell.

Steve

When I was running my projects under BOINC 6.6.36 my floating point benchmark was ~1744 MIPS. With GR 5.10.30 my FP benchmark result is ~1145 MIPS. Any thoughts on this?

Steve