I have tried now for two days to download new WU's anyone any idea how i can get some all i get from the server is "11/28/05 17:46:05|rosetta@home|Sending scheduler request to https://boinc.bakerlab.org/rosetta_cgi/cgi
11/28/05 17:46:05|rosetta@home|Reason: Requested by user
11/28/05 17:46:05|rosetta@home|Note: not requesting new work or reporting results
11/28/05 17:46:11|rosetta@home|Scheduler request to https://boinc.bakerlab.org/rosetta_cgi/cgi succeeded
"

I'm running Boinc Ver 5.2.12

All other projects running on this computer downjload fine just have a problem getting Rosetta WU's!

---

If you have WUs from other projects running, Boinc is going into EDF-Mode (earliest deadline first) so you can`t get new work from this project for a while. You can suspend all other projects so you will get new WUs ;)

---

If you have WUs from other projects running, Boinc is going into EDF-Mode (earliest deadline first) so you can`t get new work from this project for a while. You can suspend all other projects so you will get new WUs ;)

Thanks for that all loaded up now.

---

If you've been crunching Rosetta a lot and you're crunching other projects, in order to even up the playing field with regards to resource shares, it won't download new work until the other projects have close to their allocation of time. With my settings, BOINC sometimes takes a couple of days in between downloading workunits.

---

If you have WUs from other projects running, Boinc is going into EDF-Mode (earliest deadline first) so you can`t get new work from this project for a while. You can suspend all other projects so you will get new WUs ;)

Thanks for that all loaded up now.

That's one of the things that takes some getting used to with Boinc. It really does work the best if you just leave it alone and let it get on with it. I run two projects: R@H and Predictor, but Predictor is intended to be a backup in case R@H has to go offline. So I have the resource shares at 95% and 5% respectively.

What I've seen happening is that Predictor d/l a good number of WU's, and then crunches them at much more than 5% share, because of the shorter deadline (under 2 weeks vs the 4 weeks of R@H). Right now, it's just not getting any work at all for Predictor, but it will eventually.

If you go get BoincView from http://boincview.amanheis.de/ it shows a couple of extra columns that I can't get Boinc manager to show, one of which is "long term debt", aka LTD. That's a measure of how much over or under a project is in it's compute share. Right now, one of my systems has Predictor at about -21,000 LTD, but that figure is slowly approaching zero as R@H WU's get crunched.

Onces it gets to zero (or somewhere above), it'll d/l more Predictor WU's and the cycle will repeat.

P.S. you might want to either change the color of your stats display or put an image behind them. White text on this pale grey background is completely unreadable. :)

---

Onces it gets to zero (or somewhere above), it'll d/l more Predictor WU's and the cycle will repeat.

The cutoff is the negative of the task switch time. (how many seconds you let one task run before switching to another). If the applications are all wroking in round robin, the Long Term Debts will all fall within +/- the task switch time.

---

BOINC WIKI