Waste of resources.

[Conan]

This work unit keeps failing with "Process got Signal 11", over and over.
I crunched it for 65 Days (over 5.6 Million Seconds) before it failed and many tens of people have also had this same error.
I think it should be retired.

Conan

[Michael H.W. Weber]

This work unit keeps failing with "Process got Signal 11", over and over.
I crunched it for 65 Days (over 5.6 Million Seconds) before it failed and many tens of people have also had this same error.
I think it should be retired.

Thank you for reporting this to us.

Michael.

[klknoles]

I'm a casual BOINC user who loves to sign up to tons of projects and did a double take for the first time RNA World starting crunching since the deadline was incredibly close for the expected runtime, at first thinking the deadline was actually next year. Nope. This year, and so soon after starting with an expected runtime of 648 hours. The time required has been accurate after over 250 hours of work and I had no expectations of some kind of sudden acceleration so it was looking as if this WU was doomed from the start. It's almost as if this project's largest WUs and their so-immediate deadlines were created for CPUs that won't exist for a decade, and my 3.1GHz i5 with 32 gigs of RAM should be more than sufficient for any work BOINC can throw at it.

Yes, I know it is possible to request extensions in these forums, but this was rendered moot when I briefly suspended activity in BOINC and unlike the other projects, RNA World plummeted down to about 2.9% complete for what I think was over 30%. (but curiously not 0%) It is now creeping back up, but will obviously not make it, nor will I let it. I'm aborting the WU and will be suspending RNA World, only bringing it out of the coma once or twice a year to see if any changes have been made to make this a viable project. I guess it is still described as a Beta, but yet, without realistic deadlines and a way to resume work from roughly the point of suspension, RNA World simply doesn't work for me. I'm not going to bother trying to set prefs for smaller WU's, but instead hold out for the project to improve to the point at which it runs as smoothly as climate prediction.net. Now there's a long runtime project with generous deadlines and flawless resumability!

[Coleslaw]

klknoles, perhaps you don't watch the CPDN forums. They have problems all the time. So, I hope you don't want this project to run "as smooth as" they do. Also, you may want to look at the other apps here since they are working on a version that uses Virtualbox. This should cut down on the loss of time when suspending. However, I too have found the deadlines to be unreasonable. It further irritates me when work units sit on 100% for a week or more. I would say the term BETA is absolutely correct at this project and probably wont have that title removed any time in the near future.

[Dunuin]

The new VM-WUs is not only supporting checkpointing, it also extends the deadline automatically.

[klknoles]

klknoles, perhaps you don't watch the CPDN forums. They have problems all the time. So, I hope you don't want this project to run "as smooth as" they do. Also, you may want to look at the other apps here since they are working on a version that uses Virtualbox. This should cut down on the loss of time when suspending. However, I too have found the deadlines to be unreasonable. It further irritates me when work units sit on 100% for a week or more. I would say the term BETA is absolutely correct at this project and probably wont have that title removed any time in the near future.

I haven't read the CPDN forums since its solid performance has left me not wanting to find out what's wrong (hence my presence here.) I can only speak from my personal experience with CPDN and it's had no problems for me with it being an example with very long runtimes, generous deadlines, and apparently a "it just works" experience in terms of resumability, all of which should serve as a model for RNA World. That said, one of these days CPDN may crash and burn for me, but until then...

As I said, I sign up for tons of projects, including yoyo@home which seems to work fine, but for the time being RNA World has disinterested me with a bad first impression. It may be a chicken & egg problem with drawing in users and getting the project to a more viable status. Each may beget the other.

[Michael H.W. Weber]

No, CPDN will certainly not serve as a model for our purposes, because it does not have to deal with the problems, we have to deal with. And I think we dealt successfully: checkpointing is currently in (a very promising) testing phase.

Michael.