Bl00dy stuck units

Just went through my servers and found about 12 stuck units:mad:

What a waste of crunching time, some had been going for over 3 hundred hours:eek:

thats the problem with rosetta and a main reason why Ive gone 100% simap…

I have remotes I cant touch and the users dont know what to do… so its pointless,

come to the simap juggy :wink: :stuck_out_tongue:

unlucky tho! :frowning:

I have not seen a stuck one since the upgrade to 482 Rosetta.

You realy have to check round once in a while with rosetta as stuck units still happen to me. restarting boinc manager sometimes gets them going. Also had some that are running but give no percentage progress or CPU time. :eek:

Finding this all the time on my ltsp nodes, but main server never has the problem.

I’m going to aim to get to some sort of round number and then change projects :frowning:

SIMAP seems fine. I’m now running 75% on an Alpha project (QMC@Home) and that seems more stable than Rosetta! :eek:

Never had a problem with stuck units :shrug:

Must be doing something right :smiley:

Well I just went through my servers and agai8n I have about 40 stuck units. I’m going to carry on in rosetta for another 2 weeks and if things don’t improve I’ll be moving back to Seti:(

Me neither :confused:

Strange, 348 hours one had been running for and still on 1.00% completed:confused:

Wont dont you try SIMAP?
Thats been stable as anything throughtout me running it.:slight_smile:

Maybe, I’ll opbviously try the one I feel most strongly about.

Let we see how Rosetta goes over the next few weeks.

Thats fair enough. Also your near 100,000 points I see :wink:

Predictor has also managed to pick up it’s act again and seems fine :slight_smile:

I have zero problems with Linux runing Rosetta. The return results has to be kick once in a while but no hung units and only had one end early with an computation error.

It is really frustrating becuase I haven’t been able to baby sit them since last week thursday. Now I’ve wasted the last couple of days crunching:(

Juggy have you turned the user preferences setting down to 2 hours on WUs. This seems to help the 1% problem.

http://boinc.bakerlab.org/rosetta/forum_thread.php?id=1274

Are you referring to the “Target CPU runtime”?

yes that is the one.

Had it set at 2 hours, didn’t make a difference.

I’ve had issues with stuck units also. One way I’ve found of minimizing the task of herding my collection of cats is to use the “View Computers” link on the Rosetta “My Account” screen. Any computers that haven’t reported in during the last 24 hrs go on my list to have a look-see at…