[RegCNET] Some puzzles~

[Yanlb]

Dear RegCMers,
I am encountering some puzzles and expecting your help. Any suggestion is welcome.

1,
I am running the RegCM-4dust. It will halt for some clew such as
++++++++++++
 process            2 of           32
 process            4 of           32
 process            7 of           32
p7_1005:  p4_error: net_recv read:  probable EOF on socket: 1
++++++++++++
I modified the time step again and again, for example
ds=50,dt=150,450
ds=50,dt=100,300
ds=50,dt=90,270
It seems that the less time step, the more running time.
Unfortunately, It will halt still.
I want to know ds=50,dt=50,150 can be the running configuration?
2,
When the model halted like above reason, I restarted it and the running had been complete.
I want to know why the restart can complete running? The same equation or the same calculating methods...

Thank you for paying attention to my questions.
Best regards,
Libin Yan

2009-04-06 



――――――――――――――――――――――――――――――――――――――
Libin Yan, PhD candidate
Climate and Environment Modeling
Institute of Earth Environment, Chinese Academy of Sciences (IEECAS)
10th Fenghui South Road, High-Tech Zone, Xi’an, 710075, Shaanxi, China
Mobile: 86-15902909106   Tel: 86-29-88327411   Fax: 86-29-88320456
Email: yanlibin at ieecas.cn
Labpage: http://tibet.ieecas.cn/
Personal: http://tibet.ieecas.cn/ylb.htm
――――――――――――――――――――――――――――――――――――――
 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.ictp.it/pipermail/regcnet/attachments/20090406/d9f3a331/attachment-0002.html>