climateprediction.net (CPDN) home page
Thread 'model save times'

Thread 'model save times'

Questions and Answers : Unix/Linux : model save times
Message board moderation

To post messages, you must log in.

AuthorMessage
old_user3682

Send message
Joined: 30 Aug 04
Posts: 24
Credit: 250,709
RAC: 0
Message 5153 - Posted: 9 Oct 2004, 6:24:53 UTC

Hi .

one of my machines has taken to canning itself now and then just recently , This means that every time it dies the model restarts but i have lost about 3 days work .

My question is why was the save interval set to this longer a time period ? , would it not save a lot of computing time/power if the save interval was 1 day or even 12 hours that way when a system does have a glitch you dont loose quite so much time ..

Pete

ID: 5153 · Report as offensive     Reply Quote
old_user1764

Send message
Joined: 26 Aug 04
Posts: 15
Credit: 1,320,185
RAC: 0
Message 5209 - Posted: 11 Oct 2004, 0:09:13 UTC

It is set to save every 3 days of model time. 3 days at every half hour equals 3 time 48 equals 144 time steps. On a fairly slow machine at 10 per time slice that is only 24 minutes.
ID: 5209 · Report as offensive     Reply Quote
old_user3682

Send message
Joined: 30 Aug 04
Posts: 24
Credit: 250,709
RAC: 0
Message 5215 - Posted: 11 Oct 2004, 5:38:11 UTC - in response to Message 5209.  

> It is set to save every 3 days of model time. 3 days at every half hour equals
> 3 time 48 equals 144 time steps. On a fairly slow machine at 10 per time slice
> that is only 24 minutes.
>
>Hummmmmm ok now i am miffed think i see where this is heading , I'll get my head round it yet thanks ..i was getting virtual and actual mixed i think ..

Pete.

ID: 5215 · Report as offensive     Reply Quote

Questions and Answers : Unix/Linux : model save times

©2024 cpdn.org