Not only is it impractical to back up1,600+ Megs of data, not all the files care to copy over for some reason.

I\'m puzzled. Admittedly there\'s a lot of data, but are you saying it is impracticable because of a lack of space, or other reasons? It takes a minute of two, but it is a lot of data. As for not being able to copy all the files, I\'m afraid it does sound as if you have not closed BOINC down completely first. Something must be locking the files. To close BOINC you must use File/Exit if in BOINC Manager; just closing the window does not close BOINC.

As for your other comments, I think you might be confusing BOINC and the application. The size is accounted for by the data files generated in climateprediction; they are zipped when it is appropriate, but nothing can be done about the amount of data when modelling the world\'s climate in the way it is being done here. It isn\'t really like your emails. And yes, the program does have its own backup system built in, which is why it may sometimes go back and repeat work, but the backup we are talking about here is a precaution in case that does not suffice.

Data backup is routine in the commercial world and the methods used depend on the application in question. Most wise data managers try to make sure that they have a backup of all their files.

You need to make a backup of the entire BOINC folder to \"restore\" after a failure.
And you MUST suspend BOINC first, so the many files in cp don\'t get out of sync during the long copy process.
<a href=\"http://www.climateprediction.net/board/viewtopic.php?t=2377\"> This</a> thread discusses backups, with links to several other threads.

Keep in mind that the end of phase uploads in sulphur models are big; the first is over 8 Megs, which may take a while on dialup. And with version 4.19, ALL the big files get uploaded in one go at the end.

You write that you did a system restore and lost everything. How can this be? Was BOINC and CPDN running when you took the backup? If so, this would explain it. Backup cannot save all files at the same time, so when a restore is done, files will be out of synch.

I make a regular backup of the entire BOINC folder (while BOINC is down) - about once a week. If something ever happens - and it did happen once - then I can just go back and run it from there.

I run a nightly scheduled defrag (using Diskeeper with smart I/O enabled), and it never interfered with BOINC or CPDN. Windows\' own defragger may be less tactful.

You need to make a backup of the entire BOINC folder to \"restore\" after a failure.
And you MUST suspend BOINC first, so the many files in cp don\'t get out of sync during the long copy process.
<a href=\"http://www.climateprediction.net/board/viewtopic.php?t=2377\"> This</a> thread discusses backups, with links to several other threads.

Keep in mind that the end of phase uploads in sulphur models are big; the first is over 8 Megs, which may take a while on dialup. And with version 4.19, ALL the big files get uploaded in one go at the end.

It has been highly suggested to shut down BOINC and all the processing before doing a defrag. Since you are moving files that are partially in memory at this time, it can lose where the data is on the drive and fail the WU.