Cheers! I upgraded BOINC manager from 5.8.8 to 5.8.11 and noticed I receive the following messages from climateprediction.net:

2/13/2007 9:59:05 PM|climateprediction.net|Task hadcm3ohc_2a5h_05654828_1 exited with zero status but no \'finished\' file
2/13/2007 9:59:05 PM|climateprediction.net|If this happens repeatedly you may need to reset the project.

This happens with or without the screensaver activated. When I uninstall 5.8.11 and re-install 5.8.8, all seems fine even with the screensaver active.

Any thoughts or should I just stay with 5.8.8 \'till further notice?

Lee

The message is usually benign, and is probably triggered by Windows syncing the clock from a net time server.
It doesn\'t happen to all computers, and even then not all of the time.

---

Backups: Here

Hi,

I have the same problem, and also recently updated to 5.8.11. But the point is that I get the error every time boinc switches between my 2 projects, cpdn and rosetta.

On the other hand, everything seems to advance as it should be, I mean cpdn %, time and date, trickles and so on. Is there something else to look out for in order to identify a problem?

Thanks,
todor

There\'s some ideas in Pooh Bear\'s post here, and another here.

Hi,

I have the same problem, and also recently updated to 5.8.11. But the point is that I get the error every time boinc switches between my 2 projects, cpdn and rosetta.

On the other hand, everything seems to advance as it should be, I mean cpdn %, time and date, trickles and so on. Is there something else to look out for in order to identify a problem?

Thanks,
todor

You are right on here. A bug has been squashed that would cause the client to display the \"no finished file\" message for no good reason when switching tasks. Look for the fix in 5.8.13 and later clients.

---

BOINC WIKI

BOINCing since 2002/12/8

I also upgraded to 5.8.11 and got the same \"no \'finished\'file message, followed by \"If this happens repeatedly you may need to reset the project. I also noticed that there were no recent trickle uploads.

Two other projects, Einstein and SETI, displayed the same message, and there were no uploaded work units. My remaining project, Rosetta, didn\'t display the messages, and uploads were occurring normally.

I then reset the Einstein and SETI projects, which immediately downloaded new software and new work. Unfortunately the problem has persisted, and with SETI, I\'m getting a new message, \"Restarting 24no03aa.2465.9634.642322.3.83_0 - message timeout.\" Only Rosetta continues to upload finished work.

A few minutes after resetting the other projects, I reset Climate Prediction as well, and I watched 600+ hours of work dissapear from my task pane. Since then, I\'ve received no new work from Climate Prediction.

Hi cwc

If I hover my mouse over the boinc Reset button, it says \'Deletes all files and tasks associated with this project and fetches new tasks\'.

On the other hand, if someone asked me to \'reset the table\' because it wasn\'t laid properly for the meal, I wouldn\'t expect all the crockery and cutlery to vanish into thin air. It\'s not immediately obvious that Reset means losing the lot.

Have you by chance got a backup of your models that you can restore? If not, I\'m afraid there\'s no way to get crashed or aborted models back.

Your two models trickled on 14 Feb, so maybe there was nothing wrong?

To find out about how to back up the boinc folder, go to the READMEs in my sig. In the one about crashes, item #1 by Les explains an easy method. There\'s also an item about the \'Zero status\' message.

Assuming you still haven\'t received replacement climate models, if the button in the Projects tab says \'Allow new work/tasks\', you\'ll need to click it once.

---

Cpdn news

Hi mo.v

I started into the 4 project READMEs that you recommended. After reading about backups, I checked and found, unfortunately, that I didn\'t have any.

Although I still have more reading to do, I wanted to get the project started again, so in an attempt to get new work, I first toggled the \"no new work\" button on and off. That didn\'t help, so I clicked the update button. That didn\'t help, so I reset the project again. There were more download messages this time than for my prior reset operation, and this time I got new work.

I received two new work units, and I suspended everything else to force them to start. At this moment, they appear to be running OK, and both are now into December, 1920.

I\'m waiting now to see if I get a successful trickle upload. That\'s what lead me to reset the project in the first place.

Regarding my lost work units — is it practical to re-construct an incomplete work unit from the data already uploaded? Adding that ability to the project could be a big help in situations such as mine.

Thanks for your help!

cwc

Carl\'s plan has been to have crashed models that have reached the 40, 80 or 120 year restart dumps finished off on other computers. When he\'ll have time for this is another question because there are only 3 programmers at cpdn in Oxford.

---

Cpdn news

cwc said:

... At this moment, they appear to be running OK, and both are now into December, 1920.

I\'m waiting now to see if I get a successful trickle upload. That\'s what lead me to reset the project in the first place.

Trickles are created at the END of each model year. The models start in December 1920, so as to have 1921 be a complete year. The first trickle will be at 0.67%, at the end of 1921.

Lots of patience are needed when running these models.

A model only exists as a \"work in progress\" on a person\'s computer. The data that\'s uploaded each year is only a few statistics about a given model. It\'s not enough to recreate the entire model.
There is a \"restart dump\" uploaded every 40 model years, that will at some time in the future, when a program is written to allow it, be able to be used to start a model from that point in time.

---

Backups: Here

After resetting the project, I\'m now into the year 1922 with two new work units, and maybe I don\'t know what to look for, but I haven\'t seen any indication of a trickle upload. Here are all of the relevant messages since the reset. All took place on 2/20/2007:

07:44:10 AM||Running CPU benchmarks
07:44:10 AM||Suspending computation - running CPU benchmarks
07:44:12 AM|climateprediction.net|Task hadcm3pbb_bqjo_05821695_0 exited with zero status but no \'finished\' file
07:44:12 AM|climateprediction.net|If this happens repeatedly you may need to reset the project.
07:44:12 AM|climateprediction.net|Task hadcm3pbb_aflv_05760862_1 exited with zero status but no \'finished\' file
07:44:12 AM|climateprediction.net|If this happens repeatedly you may need to reset the project.
07:45:12 AM||Benchmark results:
07:45:12 AM|| Number of CPUs: 2
07:45:12 AM|| 1433 floating point MIPS (Whetstone) per CPU
07:45:12 AM|| 2212 integer MIPS (Dhrystone) per CPU
07:45:13 AM||Resuming computation
07:35:20 PM||Restarting hadcm3pbb_aflv_05760862_1 - message timeout
07:35:20 PM|climateprediction.net|Restarting task hadcm3pbb_aflv_05760862_1 using hadcm3 version 515
07:50:20 PM||Restarting hadcm3pbb_aflv_05760862_1 - message timeout
08:04:57 PM||Restarting hadcm3pbb_aflv_05760862_1 - message timeout
08:20:15 PM||Restarting hadcm3pbb_aflv_05760862_1 - message timeout
08:56:28 PM||Restarting hadcm3pbb_bqjo_05821695_0 - message timeout
08:56:28 PM|climateprediction.net|Restarting task hadcm3pbb_bqjo_05821695_0 using hadcm3 version 515
09:10:53 PM||Restarting hadcm3pbb_bqjo_05821695_0 - message timeout

Trickles are Model-yearly, about 4Dec; every 25920 Time Steps.

Your first Trickle will happen in Dec 1922. It will be credited in the next daily cron job.

---

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

Those messages telling you to reset, DO NOT RESET! These are warning messages, and they are working on the bug fix to not have that message happen less often, or better yet, when it is truely needed.

Right now lots of things can set that warning off. Including the screen saver, antivirus software, disc scans / defrags, etc. I could name probably a few more, but you get the idea.

Just let it alone, it will run and do it\'s job.

---