Sounds like he's got a handle on the problem, and a direction to go to clear it up. Now all he needs is time ........

---

What version of BOINC?

5.2.13 with the Trux optimised BOINC.

The 3 WUs at "ready to run" have a reporting deadline of 24th April, as does the one currently being crunched. The latter will report in just over 1 hour (about 9.00 am UTC).

I am trying out mmciastro's suggestion, and have "suspended new work" in the BOINC Manager Projects tab. My aim is to run out of cache and then again "allow new work". Failing that then I will detach and reattach!

---

It's good to be back amongst friends and colleagues

What version of BOINC?

If it is 4.35 to 5.2.x, it will enter EDF if the deadline is withing 2* the connect interval. It will enter NWF if it cannot make deadline within 90% of deadline. It will also enter NWF if it cannot make deadline within 90% of deadline if a fulltime result is added to the system at the resource share of the next project to be allowed to fetch work.

If it is 5.3.x, there is some missing code, and work could be returned late. (going to 100% of deadline on a routine basis is a recipe for late work occasionaly).

If it is 5.4.x (not released yet), it has a compuatation deadline that is earlier than the report deadline. The computation deadline is report deadline - (1 day + connect every X + project switch time), and the host will enter EDF if a result cannot be made with 90% of that if the next project allowed to fetch work is allowed to have work at its resource share.

---

BOINC WIKI

OK, I see you're running the Trux optimized boinc, and Crunch3R optimized application. Perhaps Trux/Crunch3r can answer your question. Me, I'd wait till they're done, and if new work weren't gotten by then, I'd detach/reattach and reload everything. Maybe the developers of the software you're using have experienced this before. It's not "official" boinc, or Seti software, and that's what I'm familiar with.

That puter currently has 5 result it's crunching, each on take roughly 20000 seconds, that's 100,000 seconds of actual work. there are 3600 seconds in an hour, 86,400 seconds in a day, this means you have 1 1/4 days worth of work. They all have a deadline of the 24th of April, you should have no trouble completing these by the deadline.

Are you attached to any other projects? If so, which ones, what is the total estimated completion times for those projects, and what are the deadlines.

No I am crunching SETI only, and have my General Preferences set to 3 days cache

---

It's good to be back amongst friends and colleagues

That puter currently has 5 result it's crunching, each on take roughly 20000 seconds, that's 100,000 seconds of actual work. there are 3600 seconds in an hour, 86,400 seconds in a day, this means you have 1 1/4 days worth of work. They all have a deadline of the 24th of April, you should have no trouble completing these by the deadline.

Are you attached to any other projects? If so, which ones, what is the total estimated completion times for those projects, and what are the deadlines.

Try a manual update of the 98se machine. Go to the project tab of the manager and highlight the project then press update. See if it starts working nomal after that.

Thanks for the suggestion.

Unfortunately I have tried that one several times without any WUs downloading. I have even rebooted the laptop and restarted BOINC without success.

I have just now pressed the Update tab under BOINC Manager's Projects tab. The Messages tab reports -

Reason: Requested by user
(not requesting new work or reporting results)
Scheduler request to http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi succeeded

I am not able to cut and paste the Messages tab report as I am using another PC for this reply!

---

It's good to be back amongst friends and colleagues

Try a manual update of the 98se machine. Go to the project tab of the manager and highlight the project then press update. See if it starts working nomal after that.

---

98SE XP2500+ @ 2.1 GHz Boinc v5.8.8

And God said"Let there be light."But then the program crashed because he was trying to access the 'light' property of a NULL universe pointer.

I have a small 4 PC farm, and I altered the General Preferences 3 days ago. The reason was to reduce my cache from 4 days to 3 days via the "connect to network about every ...".

Locally BOINC Manager on each PC is set on <Commands/Network activity based on preferences>.

After a day's crunching, when my PCs did not collect new WUs, the system started downloading new WUs for each PC. That is it did on 3 machines OK, but my old Win98SE Laptop is crunching WUs, sending off results and not collecting new WUs to maintain the cache at the new level!

Why is this machine different to the others, and what can I do to correct it?

NOTE: In BOINC Manager's Projects tab I have not activated the "No new work" buton. It still reads that title, so I should be downloading OK!

I was unable to specifically find anything on BOINC Wiki to answer this question

---

It's good to be back amongst friends and colleagues