totto73, there was an issue with the original set of se* WU's, and thus, they were aborted as soon as it was discovered. What you should notice though, is that you'll have been given your points/time.

The error was then fixed and the WU's sent out again - this time with 'erle' within the name.

This is the status at the moment.

BETA_ erlc_ a028_ pe0000_ 3-- Saturn In Progress 28.01.10 06:42:30 01.02.10 06:42:30 0.00 0.0 / 0.0
BETA_ erlc_ a028_ se0000_ 2-- Saturn Server Aborted 27.01.10 16:26:58 27.01.10 21:34:23 0.00 0.0 / 0.0
BETA_ erlc_ a137_ pe0000_ 1-- Saturn In Progress 27.01.10 16:12:23 06.02.10 16:12:23 0.00 0.0 / 0.0
BETA_ erlc_ a005_ pe0000_ 0-- Mars In Progress 27.01.10 15:39:11 06.02.10 15:39:11 0.00 0.0 / 0.0

I checked and they are being crunched one is nearly done. I just forced start the two others. But they lost too much time in my buffer queue so I suppose that I will be late and validation is unlikely even if I get the points. To get even a bronze badge will be hard.

I wonder if it is possible to have in the config file some kind of instruction that when a beta comes it takes priority on running WU's. One of the running WU is put into Wait to run state and the Beta one becomes Running state. It would be a nice feauture for those who cannot sleep in front of their CPU's with a webcam looking at the screen and that when it reads beta rings an alert siren.

Hypernova,

No, there's no config for that. Back when WCG gave 2 day deadlines on beta tests, these jobs would jump the queue by themselves pretty quick... but then there were these surprise jobs of 100 hours causing many redundant copies to go out and even more too lates.

But mostly, it's members creating the problem. Each want piles of beta tasks, so they play the buffer game. If everyone would just stick with the client default of 0.3 days cache (or whatever it is these days), each client, many more clients would fetch a job in waiting of the current one finishing and the beta test would be done in the shortest possible time.

So, Beta is as it stands hand on and then if you run them there's still a grace period for when you were not there. Want Gold before 2019, you'll indeed need to stay eyes wide before the VDU.

edit: address added

Please remember that the "se" version of the work units run quicker than the "pe" versions. This is due to the calculations needing to be done in the "pe" types.

As for writing the trj and rst file, these need to be written and read as it is part of the dynamics loop. These files are written for the purpose of the calculations needed later in the dynamics loop. If these files were not written, then the total system energy would be completely wrong and the calculations done would not be correct.

When the production work units are released we will be reducing the output to result.out which writes throughout the course of a work unit. This log file is very helpful in figuring out a situation, but is not needed for the researchers.

Hope this helps,
-Uplinger

Question? I had 6 completed WU of the inital "bad" group that were status PV at the time of your post. Shortly after the status changed to Too Late with no credits/Hours posted. Should they have received credit?

Result Name Ver Status Sent Return CPU Claim/Grant
BETA_ erlc_ a209_ se0000_ 1-- 614 Too Late 1/27/10 16:52:34 1/27/10 23:01:39 3.19 54.4 / 0.0
BETA_ erlc_ a191_ se0000_ 2-- 614 Too Late 1/27/10 16:50:08 1/27/10 22:56:33 3.20 60.2 / 0.0
BETA_ erlc_ a176_ se0000_ 2-- 614 Too Late 1/27/10 16:47:51 1/27/10 22:51:09 3.25 61.5 / 0.0
BETA_ erlc_ a164_ se0000_ 0-- 614 Too Late 1/27/10 16:46:28 1/27/10 22:51:09 3.27 61.7 / 0.0
BETA_ erlc_ a156_ se0000_ 2-- 614 Too Late 1/27/10 16:45:28 1/27/10 22:51:09 3.25 61.3 / 0.0
BETA_ erlc_ a155_ se0000_ 0-- 614 Too Late 1/27/10 16:45:28 1/27/10 22:51:09 3.25 61.3 / 0.0