All of my necessary files downloaded ok and the work unit got up to 33.33% completed. I started winamp media player and all of a sudden the work unit seemed to be going in reverse

The status is running and according to taskmgr.exe it is using 98 percent of my processor. Yet, It has failed to make any progress and the time it takes to complete continues to go up with no progress made. What is going On? I tried to restart my computer, but to no avail.

9/26/2005 10:47:15 PM|LHC@home|Sending scheduler request to http://lhcathome-sched1.cern.ch/scheduler/cgi
9/26/2005 10:47:15 PM|LHC@home|Requesting 8640 seconds of work, returning 0 results
9/26/2005 10:47:15 PM|rosetta@home|Pausing result 1btn__abrelax_no_cst_01695_0 (removed from memory)
9/26/2005 10:47:17 PM||request_reschedule_cpus: project op
9/26/2005 10:47:18 PM|LHC@home|Scheduler request to http://lhcathome-sched1.cern.ch/scheduler/cgi succeeded
9/26/2005 10:47:18 PM|LHC@home|No work from project
9/26/2005 10:47:19 PM|LHC@home|Deferring communication with project for 59 seconds
9/26/2005 10:47:21 PM||Suspending computation and network activity - running CPU benchmarks
9/26/2005 10:47:23 PM||Running CPU benchmarks
9/26/2005 10:48:20 PM||Benchmark results:
9/26/2005 10:48:20 PM|| Number of CPUs: 1
9/26/2005 10:48:20 PM|| 1135 double precision MIPS (Whetstone) per CPU
9/26/2005 10:48:20 PM|| 2168 integer MIPS (Dhrystone) per CPU
9/26/2005 10:48:20 PM||Finished CPU benchmarks
9/26/2005 10:48:20 PM||Resuming computation and network activity
9/26/2005 10:48:20 PM||request_reschedule_cpus: Resuming activities
9/26/2005 10:48:36 PM||request_reschedule_cpus: project op
9/26/2005 10:48:36 PM|rosetta@home|Restarting result 1btn__abrelax_no_cst_01695_0 using rosetta version 4.77

---

I don't have an answer, but I might be having this problem on my first WU also. It got to 91.6% complete with straight computation. Then (and it was over a half hour ago, it has remained stuck there. If BOINC had switched to another project (it had not) I have things set to keep the WUs in memory, but it's been running straight.

Task Manager also shows about 98% CPU time being utilized by the rosetta process, and it varies between 107 Kbytes of RAM and 140 Kbytes of RAM (it cycles between these values while the process is running).

I can leave it running through the night and see if it finishes.

---

OK, my WU just went up, so I gather it hit a part of the WU which was computation intensive.

I'm not sure about your case, except from what was already said in another question a reboot will cause a restart as there are no save points. Perhaps someone else has a better idea and can help you, if it's not making any progress at all. If it started making some again; it might help to leave it running as I had. Odd that it took almost an hour to complete the last 9%, when the first 91% completed in 1 hr and 45 mins.

---

<i>a reboot will cause a restart as there are no save points</i>
checkpoint exists

---

---

Well, I left my computer on through the night (7 hours) and when I came back, only one work was copleted. I think I am having the same problem. It took an hour to complete the last 10 percent. The progress on the unit is now 100 percent however the status is still running. Why is this?

I have a total of two other work units from Rosetta@home. Why is it that once the 1st work unit completed, it didn't automatically go to the next work unit that had a status of "ready to run?"

6 hours of comp. time was wasted because of this.......

---

Depending on how many projects you're signed up for, and what the resource shares are, etc; the BOINC client will go to whatever it has to accomplish. This might not be another rosetta WU. If it's processing something else after it finishes, that's normal and I've seen it times with other projects as well. Now if BOINC isn't moving to any other WU from any project though there are WUs to be processed, that might be something to look into.

Oh, and as to the checkpoints, I was getting my information from this question here

https://boinc.bakerlab.org/rosetta/forum_thread.php?id=38

Perhaps a newer version of the software has changed this. That said, the increments can include some fairly lengthy run times (such as we've been seeing here at the end of the WU crunches); so even with check points if a unit is restarted it would be rolled back to the state it was in at the last checkpoint.

---

You are getting an unhandled exception error, look here


Exit status 1 (0x1)
Computer ID 3726
Report deadline 25 Oct 2005 1:43:40
CPU time 13692.59375
stderr out 4.45
Incorrect function. (0x1) - exit code 1 (0x1)



***UNHANDLED EXCEPTION****
Reason: Access Violation (0xc0000005) at address 0x0048E868 write attempt to address 0x0A5CF440

Exiting...

---

BOINC Wiki

Now that you say it, my work unit also completed. I am not sure how long it took to complete the last ten percent (over 2 hours), but the first 90 percent completed in like 2 hours......I think there is a bug in the unit causing the huge fluctuation in processor time to complete a unit

---