When the program encounters a problem, it rewinds 1 day and tries again.
If it still has a problem, it rewinds 1 month, and tries again.
If it still has a problem, it rewinds 1 year, and tries again.
If it STILL has a problem, it gives up, and issues an error message.

This appears to be what it\'s doing with your current model. There may be something wrong with the combination of values in the parameters used, and it\'s trying to adjust. Not all combinations are valid, but the only way to find the \'bad\' sets is to run them.

Let it continue, at least for a few hours. Post again if it is still doing it, with the dates returned to each time if possible. This may be valuable info for the programmers.

Do you have some zero length *c10 files in the directory projects/climateprediction.net/hadcm3lbm_9gty_05210985/dataout?

It sounds like you\'re experiencing the same model rewind problem that a few people (including myself) have been hit with recently, and the zero length files are one of the symptoms. If that\'s the case the only way to get out of the loop is by manually aborting the result and downloading a new one.
____________
"The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer