Looks like cleaning-out started here also. I had been away for several days and the computers had lost their "trustworthy" status. Yesterday, the scheduler refused to send more than 2 workunits for empty queues. This morning, two queues were completed and empty even though two remained connected. They were set to FAAH only and the server said that there was no work.

I changed to "all projects" and immediately got some HPF2 to work on. It's about 24 hours worth of work at the moment. The queue is set to 2.5 days, so perhaps tomorrow I will get a larger percentage.

This was a double whammy today, with a minimal queue available and FAAH cleaned out.

I thought I had set BOINC to send other work if the primary project is not available. I guess not. However, I did not get more until I changed it to "all projects". I later checked the box. I must have turned it off earlier to force FAAH only.

When the new version of AutoDock starts, will the quorum of 2 be credited by averaging the two results or using the lowest claim? I am seeing instances of the lowest score being used instead of the average.

Project Name: FightAIDS@Home
Created: 08/06/2007 20:08:45
Name: faah2042_d237n441_x2AZC_00
Minimum Quorum: 2
Initial Replication: 3

Result Name Status Sent Time Time Due /
Return Time CPU Time (hours) Claimed/ Granted BOINC Credit
faah2042_ d237n441_ x2AZC_ 00_ 2-- Valid 08/08/2007 03:10:42 08/08/2007 10:23:42 7.12 62.2 / 60.2
faah2042_ d237n441_ x2AZC_ 00_ 1-- Valid 08/08/2007 03:01:02 08/08/2007 09:08:18 5.71 60.2 / 60.2
faah2042_ d237n441_ x2AZC_ 00_ 0-- Valid 08/08/2007 02:58:46 08/08/2007 07:04:13 4.04 73.8 / 60.2


My result is the one claiming 73.8 and returned first. Then the 60.2 was returned and became the granted credit. My claim is approximately 22.5% higher than the other. Is that enough to be considered an outlier?

I am running an Athlon64 X2 6000 at stock speed, WinXP, cc 5.10.16

Thanks, I saw that about 2 minutes after making my post. It helps to read through all of the recent threads before asking a question.

Let the jobs finish for upload and force the "ready to report" update. That sends the 'good machine' message' to release more work.

Else if not, reset project. de-tach, reattach from a test i did yesterday, retained the same device ID, so no worries in that department.

Midday was already 117 CPU years due the accilerated flush. What is gained today is paying forward or was is backward ;>)

Come launch of the FA@H update lots of work will go initially into PV.... fast machine matched with slower machine, thus expect an extra dip coming weekend or?