Resetting of cpu time to zero also happens if Boinc it's doing benchmarks. Application in progress will be removed from memory and if this is rosetta, then cpu time is zero after benchmarks. Just saw that on my linux box.

The last WU that I processed involved a BOINC benchmark running in the middle of the WU. It dumped the WU out of memory like it should and brought it back after the benchmarking was completed without losing any of the CPU time. Looks like they fixed that problem.

---

yup that happened to me the other day and it just started back where it had finished

---

Join us in Chat (see the forum) Click the Sig


Join UBT

decided to try again. suspended my other 2 projects and started rosetta. success! finished in 8+ hours. ;)

Same here, with the optimised app (4.76). Took only 2 hours and 25 minutes on my Athlon XP 3000+. Looking good.

---

*** Join BOINC@Australia today ***

With new 4.76 app 5:30h for one WU on a 3,2Ghz P4. Nice to see that without any errors but checkpoints are too high (1h) I think. It would be better after 10min.

---

With the new 4.76
1) 1h 36 min
2) 1h 30 min
3) 1h 34 min

no errors

Barton 3200+
Win2k

CC 4.32

---

with the optimized app (4.76), my next wu finished in 1hr 38min. :)

We're too efficient now! The result queue draineth! :)

---

After running a few WUs through my machines, I'm seeing a range of 4 to 4 1/2 hours per WU.

Machines are "rescued" Dell GX110s w/CPU speed of 933meg and 256meg RAM, using Win2000Pro for OS. Cannot cycle between Rosetta and other projects without Rosetta erroring out, so these machines are banging away 100% until I set "no new work".

---

Machines are "rescued" Dell GX110s w/CPU speed of 933meg and 256meg RAM, using Win2000Pro for OS. Cannot cycle between Rosetta and other projects without Rosetta erroring out, so these machines are banging away 100% until I set "no new work".

With the 1200 mhz AMD, 512 ram, WinXP PC I run Rosetta on, I've been able to cycle Rosetta with other projects with no problems at all.

---

Cannot cycle between Rosetta and other projects without Rosetta erroring out, ...

Please excuse my incomplete comment. I meant to indicate that I could not switch between BOINC projects without losing the Rosetta WU. Normal operation of the machine complete with other tasks such as Word, Excel, etc. work fine, including play of "recreational" programs.

---

It might be a memory issue seeing as you only have 256MB of RAM and 512MB is recommended.

---

512 is the reccomendation. I'm running dozens of machines with 128, and while they take a long time, they finish. The new wu's also take a lot less memory, somewhere between 50 and 100 rather than 200+.

---

With a celeron 1.8 GHz processor, 496 mg of RAM, and windows xp, it takes about 4 hours a work unit, about 4.5 hours if BOINCing with another project.

---

Mine took about 36 hours on a 1200 mhz AMD with WinXP and 512 ram.

Using Linux have only had "Computation errors" and using my Valuable Time
I could be giving to an Experiment that will send me info "Why" or "What" I can
do to eliviate this Issue.
I have nearly 800 RAM and there should not be only Errors,unless Linuxos
is not supported?
Sluger1138

---

With a celeron 1.8 GHz processor, 496 mg of RAM, and windows xp, it takes about 4 hours a work unit, about 4.5 hours if BOINCing with another project.

On further experience, it takes much longer if BOINCing because the unit keeps going back and starting over again at a previous point. The total time spent on the work unit also goes back. The temporary solution is to suspend other projects until I completely finish all Rosetta units as it takes too long otherwise. The client is stable, however, and this seems like a great project with a direction and purpose I do not see at other protein folding projects. I am confident that with more waypoints the program will be able to run smoothly with other projects.

---

On further experience, it takes much longer if BOINCing because the unit keeps going back and starting over again at a previous point. The total time spent on the work unit also goes back. The temporary solution is to suspend other projects until I completely finish all Rosetta units as it takes too long otherwise. The client is stable, however, and this seems like a great project with a direction and purpose I do not see at other protein folding projects. I am confident that with more waypoints the program will be able to run smoothly with other projects.

I had the same problem with my Macs. The solution was to extend the time for switching between projects from 60 minutes, to 120 or 180 minutes under general preferences. This way the next step is able to complete. Of course there may still be some fall back on each switch between projects, but it has less of an impact. Right now though, all my computers are crunching strictly Rosetta. I like this project.

---

Team MacNN - The best Macintosh team ever.

Mine took about 36 hours on a 1200 mhz AMD with WinXP and 512 ram.

Using Linux have only had "Computation errors" and using my Valuable Time
I could be giving to an Experiment that will send me info "Why" or "What" I can
do to eliviate this Issue.
I have nearly 800 RAM and there should not be only Errors,unless Linuxos
is not supported?
Sluger1138

I'll look into these errors when I get back from vacation.

Using Linux have only had "Computation errors" and using my Valuable Time
I could be giving to an Experiment that will send me info "Why" or "What" I can
do to eliviate this Issue.I have nearly 800 RAM and there should not be only Errors,unless Linuxos
is not supported?
Sluger1138

Linux is definitely supported as there are a lot of Linux boxes completing work. Try posting your detailed errors in the Linux questions and answers section. Some other Linux users may be able to help. Of course David Kim has also said he will look into it when he gets back from vacation, so help is available.

(David K, for a guy on vacation you are spending a lot of time on the forum. I am sure folks appreciate it, but don't forget to enjoy yourself some too. Wouldn't be a vacation otherwise. - David R, AKA Shaktai)

---

Team MacNN - The best Macintosh team ever.

Using Linux have only had "Computation errors"

Sluger1138,
I brought my system up on linux and tried crunching Rosetta. I haven't had a single problem with any of the five WU's I've done so far. The project definately supports linux platforms. Can you give us a little more information about your system and what's happening when you see these errors occur?

---

Mine took about 36 hours on a 1200 mhz AMD with WinXP and 512 ram.

Using Linux have only had "Computation errors" and using my Valuable Time
I could be giving to an Experiment that will send me info "Why" or "What" I can
do to eliviate this Issue.
I have nearly 800 RAM and there should not be only Errors,unless Linuxos
is not supported?
Sluger1138

Fixed Problem and am swimming alomg just fine.Just cannot exit Boinc.Plain and simple.Works great.Love the Project,will become Premiere "Experiment" shortly.
Coodos Rosetta Volunteers,Scientists and payed staff.{if there is any}.
Great W/U,Happy crunching all
Doug
Boinc Synergy

---

New WUs take much more time to finishing, between 6-9h on my P4 3,2 HT. Is there a difference with old ones (3-5h)?

---