Hey all, I'm trying to figure out why my projects keep failing - and ultimately why I'm not getting any credit for any of the tasks that do complete. Any idea what might be going on?

Thanks!

Looking at both your computers, a lot of the tasks that have failed are the short models from the 2007 batch which is dodgy. They have been removed but those that did reach computers will go out a few more times before finally being labled as failures. If you get any of these, just abort them.

There are some other work units around at the moment which also give problems on both Windows and Mac machines. At the moment Linux seems to be the best supported platform by the current batches of work.

Worth looking at the credits thread in number crunching section.

summary: Credits script is only being run manually sporadically when programming staff have time due to problems with it giving spurious results if run automatically.

Awesome. Thanks heaps for the reply.

Hi there, sorry to revive an old thread, but only noticed the problem now.

Are the explanations provided above still valid and relevant?

Noticed my Project Properties pop up states: "67 completed" and "75 failed"; the most recent failure being again a "error while computing" in December.

Also, have not seen any tasks since, with a flat line graph in the BOINC Stats tab since early December!

Thanks
Paul

Hi Meerkat

The info in this thread is too old to be relevant to current tasks/problems.

Most, if not all of your failures is
Global Worker:: CPDN process is not running, exiting
I think that this means that the monitoring part of BOINC has lost contact with the task, and then thinks that the task has stopped, so it starts the terminating process.
If so, it most likely means that something isn't being allowed enough resources, and so, by the time the monitoring process gets contact again, it has already been reported as a failure.

Not very good wording, and I'm not sure if this is the cause, but it's something like that.

And you haven't received any work for a while because there hasn't been any.
A few thousand were put up about half a day ago, but with several 10s of thousands of computers waiting, each with multiple processors, they wouldn't have lasted long.

Three things worth checking are
1 If the failure times are likely to be when you are doing something processor intensive with your machine
2. Make sure leave non GPU tasks in memory while suspended is ticked. (This under Computing preferences > Disk and memory. It doesn't stop failures but seems to reduce the percentage on many machines.
3. Make sure boinc folders are excluded from routine scans by virus checkers as the virus checker putting a lock on a file sometimes causes errors.