Same here, on a couple different tasks that then failed because it couldn't find those files.

I keep getting permanent "HTTP error" when it downloads a new file. I have now received this error twice and the project has to be aborted.
L

One PNW failed after a few seconds on a newly-rebuilt machine. All three in the Work Unit failed in, apparently, the same way. (Can't check because the server fails requests to look at any of the failures in the Work Unit.)
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/workunit.php?wuid=8172403

Probably not a machine failure (I hope!) because three PNW ran together (with a WCG task) and completed okay. (47.03 / 47.08 / 47.09 hours, running concurrently on stock i5-3550.)

---

"We have met the enemy and he is us." -- Pogo
Greetings from coastal Washington state, the scenic US Pacific Northwest.

I haven't received any new WU since June 5th, so I'm now out of work on 2 hosts.

Ven 29 jui 02:34:51 2012 | climateprediction.net | Sending scheduler request: To fetch work.
Ven 29 jui 02:34:51 2012 | climateprediction.net | Requesting new tasks for ATI
Ven 29 jui 02:34:54 2012 | climateprediction.net | Scheduler request completed: got 0 new tasks
Ven 29 jui 02:34:54 2012 | climateprediction.net | Project has no tasks available

I haven't received any new WU since June 5th, so I'm now out of work on 2 hosts.

Ven 29 jui 02:34:51 2012 | climateprediction.net | Sending scheduler request: To fetch work.
Ven 29 jui 02:34:51 2012 | climateprediction.net | Requesting new tasks for ATI
Ven 29 jui 02:34:54 2012 | climateprediction.net | Scheduler request completed: got 0 new tasks
Ven 29 jui 02:34:54 2012 | climateprediction.net | Project has no tasks available

Actually thats an ATI request. The project doesn't provide work for ATI or Nvidia. However there are only the occasional resends for CPU work requests at the moment until the scientists create more work units. There were some pnw's created about a week ago but they disappeared pretty quick (I managed to get some on 2 machines, but the rest missed out).

If you are running BOINC 7.0.25 or later there is a way to prevent it requesting ATI and/or Nvidia work (on a per-project basis) which it will never get and then backing off for an hour before it can ask again. If the BOINC server software was more up to date the project could do it as well, but they run a pretty old version. If you'd like to know how just ask.

---

BOINC blog

For the last few weeks the only WU available have been re-issues of WU that failed earlier. I've got a few of them.
Some were WU that hit broken machines on the first pass - these are working ok on my machines.
Some were misconfigured at the get-go and fail again.

When the backlog of these types of WU clears -- thanks to all you contributors --

Hoping there's funding for more :)

One PNW failed after a few seconds on a newly-rebuilt machine. All three in the Work Unit failed in, apparently, the same way. (Can't check because the server fails requests to look at any of the failures in the Work Unit.)
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/workunit.php?wuid=8172403

Probably not a machine failure (I hope!) because three PNW ran together (with a WCG task) and completed okay. (47.03 / 47.08 / 47.09 hours, running concurrently on stock i5-3550.)

---

There IS testing going on.
And Christmas will be here soon, so if you've been good ...

---

Backups: Here

And Christmas will be here soon

Ah, another time with no staff @ Oxford Uni.

I'm in no hurry since Summer is here, and it's a hot one. Once Fall starts, I will look forward to having a few thousand WU to crunch. :)

Edit: That said, running Linux means I don't get any of the EU models... so waiting is required in my case.

EU models seem to run fine on my linux box.

I just got another EU today on my Linux machine. DJStarfox, are your account settings set to receive EU's? Maybe you've inadvertently set your machine to location based preferences?

...or maybe you meant a hardware reference, such as an SSE2-less processor like an Athlon XP or Pentium III. Those definitely can't run the EU's.

Odd. I guess I am mistaken. I see now the apps are listed
http://climateapps2.oerc.ox.ac.uk/cpdnboinc/apps.php

I did check my preferences, and it's set to receive all the AM3 regional models.

I guess we're back to the empty WU queue. I subscribed to the project RSS feed, so I'll check back for work once I hear something.

The i5 laptop had three tasks, of four running, that decided to go into 'high priority' mode. When this happened no more tasks were requested by BM 6.2.12, despite the '10 days work' setting. After the fourth task completed, three tasks were left running, BM subsequently requested work and downloaded one task. Now that it has four tasks, three high priority and one normal priority, BM is still not requesting any more work despite the setting for '10 days' work.

During nearly 8 BOINC years solely on cpdn, although sight of 'high priority' is rare, I have never seen BM with this behaviour before. I presume this is BM doing as it wishes? Just curious if anyone has seen this behaviour, if is a BM version issue in 6.2.12, or a common BM behaviour?

---

I too have noted over the last weeks that BOINC Manager is more reluctant to request new tasks than it used to be. Of course, I know that there are very few CPDN tasks available just now. But, like hagar, it is the absence of requests in the event log which seems surprises me. At present, I have 4 tasks running, one of these due to finish in 16 hours, none waiting to start, no other BM apps running and my preferences set for a maximum work buffer of 9.9 days. But no requests for new work. Curious.

I see you're running BOINC 7.0.25 Lockleys. That has a change in the way the network connection parameters in your global preferences work:

• The "Computer is connected to the Internet about every" setting is now the refresh level. If you have it set to 0 BOINC won't fetch work for a project until it is starved of work.
  
• The sum of that and the "Maintain enough work for an additional" setting is now the maximum work buffer size for the project.
  

I've no idea what's preventing hagar's 6.2.12 from requesting work, but enabling work fetch debug might give some clues.

---

"The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer

Dear Thyme:

You mention that in the BM 7.0.25 if the �Computer is connected to the Internet about every� is set to 0 then the BM will wait until the machine is starved for work before requesting more. Is this the same in the 7.0.28? If so,what setting should it be set at to have it begin requesting work at the 240 hour (10 day) mark?

---

This applies to ALL BOINC 7.* versions, because 7.* is a completely new build from the ground up.

I don't know the wording of these new versions, but would have thought that they'd be different to reflect the new usage.

If it's still the old text, then try:
"Maintain enough work for an additional" to 9
and
"Computer is connected to the Internet about every" to 1

That should give you 10 days of work, then no more until your work level/load drops down to 1 day to go.

---

Backups: Here

THANKS!!!

---

Thanks Thyme and Les. That advice worked a treat. Rgds