Since going to 6.6.36 on one machine I have been receiving the following messages, so went back to 6.4.7 and everything works okay. Seems to be a scheduling problem.

My machine is a Del Dimension 3000 Pentium(R) 4 CPU 2.80GHz runnning XP service Pack 3 with all the latests updates applied.



Host Project Date Message
OFFICE Milkyway@home 6/17/2009 11:12:52 AM Message from server: No work sent
OFFICE Milkyway@home 6/17/2009 11:12:52 AM Message from server: (won't finish in time) BOINC runs 98.7% of time, computation enabled 100.0% of that
OFFICE Rectilinear Crossing Number 6/17/2009 11:13:02 AM Message from server: No work sent
OFFICE Rectilinear Crossing Number 6/17/2009 11:13:02 AM Message from server: (won't finish in time) BOINC runs 98.7% of time, computation enabled 100.0% of that
OFFICE Einstein@Home 6/17/2009 11:13:12 AM Message from server: No work sent
OFFICE Einstein@Home 6/17/2009 11:13:12 AM Message from server: (won't finish in time) BOINC runs 98.7% of time, computation enabled 100.0% of that
OFFICE Virtual Prairie 6/17/2009 11:13:33 AM Message from server: No work sent
OFFICE Virtual Prairie 6/17/2009 11:13:33 AM Message from server: (won't finish in time) BOINC runs 98.7% of time, computation enabled 100.0% of that

---

Question: Do you have any very long tasks running such as anything in the tasks tabs from Climate Prediction?

There seems to be a known bug with the scheduler when very long tasks are running. For me if I suspend CPDN, and update, tasks are downloaded normally.

No I do not have any long running jobs on the second computer that just began to faill the same way. With WUs not being downloaded because they would not finish in time.

This machine is Pentium (4) 3.00 ghZ WITH 2 GB OF memory. Also running XP service pack 3 with all the latest updates.

---

Question: Do you have any very long tasks running such as anything in the tasks tabs from Climate Prediction?

There seems to be a known bug with the scheduler when very long tasks are running. For me if I suspend CPDN, and update, tasks are downloaded normally.

I too experienced this problem after updating my 2 machines to BOINC v6.6.36. Both machines run BOINC 24/7/365 at 99.x%. Both are running CPDN workunits (of the new shorter duration requiring only ~100 CPU hrs vice 1000+, still with a 1 year deadline). One is 64-bit Vista and the other is 32-bit Vista, with dual core processors and no GPUs. I experienced the problem ONLY with Milkyway@home, running at a 5% resource share out of 10 projects attached. Seti@home Beta Test, also with a 5% share, downloaded work normally.

If as Hammeh says this is a "known bug with the scheduler," it definitely did NOT manifest at my house with BOINC versions prior to 6.6.36.

There is an active thread on this topic on the Milkyway forum that puts forward various workarounds. Personally, I reverted back to v6.6.28 and things are working fine again. I hope the BOINC development team has this as a priority issue. (Moderator please take note.)

73,
J. Whalen, N6KTC

---

This appears to be a similar/same issue to that described in thread: http://boinc.berkeley.edu/dev/forum_thread.php?id=4015 . A post in that thread suggests it is being looked into. Posts about this issue on the Rosetta@Home forum indicate that this issue is a known bug. Those posts are less confident that the matter is being addressed.

---

Here is the above-mentioned link to the similar thread properly formatted: http://boinc.berkeley.edu/dev/forum_thread.php?id=4025

---

Here is the above-mentioned link to the similar thread properly formatted: http://boinc.berkeley.edu/dev/forum_thread.php?id=4015

And here it is with the correct address ;-)

---

Computer sind nicht alles im Leben. (Kleiner Scherz)

I hope the BOINC development team has this as a priority issue.

They don't. They're busy with a whole different project that has to be done before a certain deadline. I can't say much more about that.

(Moderator please take note.)

I wouldn't know why. For everyone involved in both threads, you're way better off posting this on either the Boinc Alpha or Boinc Development email lists. That way you get in on the developers email boxes directly. Both lists need registration.

I am too busy with work for the DrugDiscovery and Hydrogen projects to completely switch over my mind to this problem, or post for you on the lists above. When you do post there, post as if you're on the forums. Don't just post a link to this or the other thread.

---

I'm seeing the same problem with 2 boxes that I upgraded to 6.6.36. Both boxes run BOINC 24x7 and do not run Climate projects

Same here....

10/07/2009 20:57:02 Milkyway@home Sending scheduler request: To fetch work.
10/07/2009 20:57:02 Milkyway@home Requesting new tasks
10/07/2009 20:57:07 Milkyway@home Scheduler request completed: got 0 new tasks
10/07/2009 20:57:07 Milkyway@home Message from server: No work sent
10/07/2009 20:57:07 Milkyway@home Message from server: (won't finish in time) BOINC runs 93.2% of time, computation enabled 100.0% of that

FYI, each task takes <67 seconds on this host ! (HD4780 in work).
They just can not be finished in time, absolutely can't...

---