I\'ve been running two slab models of late on a dual core. One has developed into an ice-world and so has been aborted. Can I reinstate a previously backed-up coupled model onto the vacant core, and, if so, how?

It is possible, although not easy - see the \'Backup and restores\' README, one of the links describes how to restore a single model from a backup. Link in my signature.

---

I'm a volunteer and my views are my own.
News and Announcements and FAQ

Thanks for the info. I think I\'ll perfect my brain surgery technique before I have a go at that!

Hi MW

I don\'t think I\'d be capable of following Pete\'s method, though it is all very clearly explained. I\'d do this:

* First check the workunit link for the backed-up coupled model to see whether anyone else\'s computer has now completed it. If it\'s now been completed by another cruncher, I\'d forget about it. If not, I\'d want to complete it myself. (If you want us to help check on this you\'ll need to give us a link to the coupled model.)

* I\'d download a new HADSM slab to keep the current survivor company on the other core, checking my CPDN project preferences first to make sure I didn\'t accidentally get a new coupled HADCM.

* I\'d attach to another project with shorter workunits and crunch some of these with the survivor suspended to even up the completion times of the two slabs. PrimeGrid and SIMAP have short WUs.

* When necessary I\'d set all the projects to No New Tasks. I\'d complete and report everything.

* I\'d then delete the contents of this BOINC folder and restore the BOINC folder containing the HADCM coupled model.

* If any model that I\'d already completed was restored alongside the HADCM, I\'d abort it. Then run the HADCM.


Anyway, that\'s a possible alternative method that avoids the selective restore.

---

Cpdn news