I\'ve just had a Windows error message saying that one of my two models (hadcm3bm_9yv6_05234357_0) has hit a problem and needs to close. The Boinc Manager says it is still running, but the CPU time measure isn\'t moving. There are no error messages in the \'messages\' tab. Help! What do I do now? The Manager had just sent a trickle-up message. Is it possible to retrieve the data from this and carry on from there? This model has had over 1,000 hours of CPU time. I\'d hate to lose it now. 22.41 BST Don\'t know what\'s happened to change things, but the model\'s running again. Panic over.

---

Is it possible to retrieve the data from this and carry on from there? This model has had over 1,000 hours of CPU time. I\'d hate to lose it now.

No.
A trickle only sends a small amount of model data to the server.
There\'s megabytes of info needed to re-start a model.
Which is why you shoul make regular backups yourself.

Just Exit from the model/program, then COPY the BOINC folder and PASTE it all somewhere else.
To restore, just do the reverse, and replace the faulty folder with the copy.