Ah, now we're talking with some tangibles. Nothing to do with your client/install, it just seems that this badge batch of results produces output much larger than the preset limits permit. Something the techs have to look at and tweak at their end.

Unfortunately, they wont be in for a few hours, so can only suggest to temporarily switch to HCC project till it blows over. Quickest is via My Grid > My Projects to apply the selection change to all your devices, if all are effected.

edit: batch of course

Lighthouse,

I have taken a look at your logs of a few of your work units on that machine. It appears that you have restarted the work unit 133 times in one case. Each time a work unit restarts for Autodock application it prints about 4000 lines of information to make sure that the system is still properly set up for the scientists.

This issue is kind of unique to the HFCC project because these work units that are running have 255 runs in them which is the most we have put in a single Autodock job at World Community Grid. I say kind of because if the other Autodock programs were to use 255 runs it would be similar, but at the moment FA@H runs less than 100 (usually around 50) per work unit.

I'm not sure what is causing your computer to restart these work units so many times. It could be anything from your computer being shut down and restarted many times in a day. To having multiple BOINC projects selected (meaning something other than World Community Grid using your computer).

Can you please post more from your messages tab? I am curious to see what is causing the restarts on your application.

Thanks,
-Uplinger

Sek,

When I was looking into the results for Lighthouse, it appeared his machine was set to Home profile which was set to standard. This allows the agent to run while computer is active. But this also removes apps from memory which is where I believe he may have more than one BOINC project attached. Don't know for sure at this moment.

-Uplinger

Not specifically. I am trying to figure out if he has multiple boinc projects selected. If so and his preferences are to switch between applications every 15 minutes, it could cause the application to stop, unload from memory, and then start again. It all depends on the logs I'm waiting on to figure out what is causing the computer to restart from checkpoint so often. I was just mentioning being attached to other BOINC projects as a possible cause to this, but without the logs it's all speculation :) And if that is the case then we will want to check how often it is told to switch between applications.

The current settings for result logs to be sent back is 10MB before compression. This is to help catch issues that may be causing the computer to restart frequently on a work unit. Again there are many ways this could happen but I'm trying to rule out one scenario. Logs will help rule out multiple scenarios...

I hope this helps, if not feel free to ask more questions :)

-Uplinger

Hi lighthouse,

As per the suspicion Post on: Nov 20, 2009 5:27:41 PM , the preventative action in your setup is indeed to tick the LAIM box. It will increase your client's crunching efficiency quite a bit by the looks of it as each time the BOINC is allowed to resume, the science will start right from the point it was paused and not from the last checkpoint.

Let us know, but think it's save to say that you can fully resume HFCC crunching.