We've pretty exhaustive discussed the dislike of VINA based sciences in the DSFL/GFAM forums and BOINC throttling, plus the workarounds [and why would most of us throttle on a desktop anyhow as it's default set for temp concern reasons of laptops/notebooks]. This issue I see not being addressed.

This beta is for another purpose than the DSFL [discover ligands sizing], i.e. were not particularly cut to find run times, but bugs...longer running ones have more chance to encounter them.

Hope this soothes the concerned minds [and somehow I made it all up without being factually told].

--//--

EDIT: A P.S. for some: Don't participate in beta if this interferes with your regular crunching... many will be happy to process them for WCG! Got 14, that got halted when the power-out lasted too long this night (over 6 hours). They just resume promptly, to include 8 on a machine that had not hibernated autonomously [something to address later]. Now all run high priority as the on_frac/active_frac were got deflated due the long outage.

Not sure if it is a real issue but currently I have 2 of these running and my hard disks are going totally mad. Never known this before with any other WU. Running Ubuntu 11.10

I agree with the consensus that these are long WUs but don't have an issue with it. However, I would be loathe to attempt a windows restart because these have very long checkpoints, for example right now I can see several 30-45minutes checkpoints and even one in the 50's!

A planned shutdown would take a while as each I wait for each long checkpoint to finish. An unplanned reboot will on multi-core machines would mean literally hours of wasted processing time which equates to wasted energy.

Despite their length, they appear to be working without incident.

@tomast, do you have boinctasks? I find it much easier to see checkpoint times than with boincmanager. For that reason alone, I would reccomend it to you.

@ littlepeaks, I have Nortons IS running and managed to get the graphics to show without a problem. Maybe because I tell it to ignore the BOINC folders and their contents.

Wow, I apparently left quite a bit out in the open for questions :)

1. General usage of the computer is asked, meaning if you want to test the shutdown issue you can, but law of averages will tell us as well as error codes recorded if the shutdown bug still exists. If you shutdown your computer without telling boinc to stop vina applications then it starts back up with a computational error ( not the expected restart from checkpoint )

2. These work units are not sizing work units like the previous dsfl beta test, they are production work units as we will encounter the windows shutdown issue more than likely with these longer work units.

3. yes only post issues with beta work units associated with this application in this thread :)

I think that addresses most of the questions/concerns. If there are others, please let me know.

Thanks,
-Uplinger