QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: Jenny on June 11, 2013, 20:02

Title: my logfile stopped which means my calculation stopped?
Post by: Jenny on June 11, 2013, 20:02

hi, everyone.

I used three computers doing parallel calculation. After three days, when the left elctrode, right electrode and bulk equivalent calculations all finished as shown in log file. The output logfile freezed at the 29th step of following calculation of transmission spectrum for one day(shown like following).

|  335  Cu   [  10.864 ,   0.234 ,  28.701 ]   11.00383   0.00383              |
|  336  Cu   [   0.639 ,   2.790 ,  28.701 ]   11.00386   0.00386              |
|  337  Cu   [   3.195 ,   2.790 ,  28.701 ]   11.00376   0.00376              |
|  338  Cu   [   5.751 ,   2.790 ,  28.701 ]   11.00322   0.00322              |
|  339  Cu   [   8.308 ,   2.790 ,  28.701 ]   11.00317   0.00317              |
|  340  Cu   [  10.864 ,   2.790 ,  28.701 ]   11.00375   0.00375              |
|  341  Cu   [   0.639 ,   5.346 ,  28.701 ]   11.00340   0.00340              |
|  342  Cu


Does it mean there is something wrong with my calculation? Or it means my calculation stops?

Thank you very much.

Jenny

Title: Re: my logfile stopped which means my calculation stopped?
Post by: kstokbro on June 12, 2013, 08:57

You did not attach the log file, so we cannot see where you are in the calculation.

If it is a heavy calculation, it might still be running but waiting for more output. On linux you can use top to see the status of the job, and in particular if it is using a lot of memory and might be swapping.

Title: Re: my logfile stopped which means my calculation stopped?
Post by: Anders Blom on June 12, 2013, 22:49

Most likely it has run out of memory, or (since this is a parallel calculation) one of the calculations on the slave nodes has terminated with an error, which can leave the master node running (or rather, waiting).

You should consider if you have enough memory for the calculation and the number of MPI processes per node.

Title: Re: my logfile stopped which means my calculation stopped?
Post by: Jenny on June 14, 2013, 00:05

Thank you for your reply. I've stopped my calculation. It seems like there is something wrong with my windows 8 system security setting.

Title: Re: my logfile stopped which means my calculation stopped?
Post by: Jenny on June 14, 2013, 00:08

Quote from: Anders Blom on June 12, 2013, 22:49

Most likely it has run out of memory, or (since this is a parallel calculation) one of the calculations on the slave nodes has terminated with an error, which can leave the master node running (or rather, waiting).

You should consider if you have enough memory for the calculation and the number of MPI processes per node.

Yes! It seems like my slave node under win 8 system stopped working which hasn't been set up consistent with my master node which is win 7. Thank you very much!