I have been running a sulpher model since last November (WinXPHome/Pentium4 machine, BOINC version 5.2.13) and it\'s over two-thirds done. A couple of days ago McAfee released a DAT file that mis-identified a bunch of .exe files as viruses and moved them under a quarantine folder. This occured during a daily scan of my computer, a scan which has been running without issue along with BOINC climate prediction and SETI for months.

One of the files moved was c:\\Program Files\\BOINC\\projects\\climateprediction.net\\sulphur_4.22_windows_intelx86.exe

After McAfee released an updated DAT file to fix the issue, I copied all the files back under the location from which they were removed (per the AV log file). I put the climateprediction.net file back under the above path and restarted BOINC but don\'t see the project listed under the Work tab.

I trie restarting BOINC a couple of time, Using the Update button from the Project tab, restarting my computer, etc but I can\'t seem to get the sulpur model to start running again. The fix may be to simply reinstall or reattach to the project but I would like to avoid possibly wiping out the 4 months of processing that I\'ve already done on the model.

Has anyone had any experience with this type of issue? Any help would be appreciated.

---

If you look at the page for the model, (in your account), you will see that BOINC has already told the server that it\'s crashed. BOINC will also have altered it\'s flag in client_state.xml which tells it what stage it is up to.

So the only way to restart that model is from a previously made backup.
You should have received a new model, but it\'s possible that program files got out of sync or corrupted. In which case, try a reset.

And exclude the BOINC folder from virus scans.

If you look at the page for the model, (in your account), you will see that BOINC has already told the server that it\'s crashed. BOINC will also have altered it\'s flag in client_state.xml which tells it what stage it is up to.

So the only way to restart that model is from a previously made backup.
You should have received a new model, but it\'s possible that program files got out of sync or corrupted. In which case, try a reset.

And exclude the BOINC folder from virus scans.

So you\'re saying that the model my machine has been working on is irretrievable unless I have a backup copy from some moment before the problem occured? Does the program make a backup of it\'s most recent good result to which I can revert? Frankly, this machine doesn\'t have personal info and I don\'t do a backup of the machine so unless BOINC or the CPDN software does an internal backup of some kind I don\'t have one.

I\'m a little confused, however, why the server would accept an out-of-date request from a backup copy but there\'s no way for me to restart the data I have on my machine that is almost certainly uncorrupted. CPDN wasn\'t executing when McAfee moved the file, it just wasn\'t able to restart once BOINC tried to hand off from SETI to CPDN. The logs show it trying to start for a couple of seconds and then nothing.

---

Most of us have started making regular backups of the BOINC folder after having experienced a crash of a ClimatePrediction WU. CPDN or BOINC do not make backups of their own - it\'s not their job.

What I am wondering is why your model crashed? Do you run other applications beside CPDN? If so, that would explain it - when trying to reload the CPDN core application, and it\'s not there anymore.

If you do not run other apps, then I cannot imagine why it crashed. It should just have continuing crunching until you stopped it, and restart normally after you restarted BOINC.

It crashed because McAfee trashed his EXEs! ... we had a general warning about McAfee yesterday at work, presumably for the same reason (not CCE exes, but simply random EXEs from different places).

---

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

Most of us have started making regular backups of the BOINC folder after having experienced a crash of a ClimatePrediction WU. CPDN or BOINC do not make backups of their own - it\'s not their job.

What I am wondering is why your model crashed? Do you run other applications beside CPDN? If so, that would explain it - when trying to reload the CPDN core application, and it\'s not there anymore.

If you do not run other apps, then I cannot imagine why it crashed. It should just have continuing crunching until you stopped it, and restart normally after you restarted BOINC.

Yes, McAfee decided to quarantine a bunch of exe files during an On-Demand scan. On-Access scanning didn\'t seem to be a problem. I have excluded database/log folders on servers but I haven\'t had any problems with McAfee for a long time and it didn\'t occur to me to exclude anything on my workstation at home.

I will make a backup going forward but I just wanted to verify that there is nothing I can do to get the current model going again before I start a new one. I thought it possible that the program could be setup in such a way that it could failback to the last good state. Just a thought. The work units for cpdn are so long that it\'s depressing to just write off the months of processing already done.

Thanks for your responses.

---

I thought it possible that the program could be setup in such a way that it could failback to the last good state.

Actually, it is programmed to do just that. If it encounters a crunching problem, it will rewind a day, then a month, then a year, trying to recover. If all that fails, it crashes and calls home.

None of that has anything to do with backups, though.

---

"We have met the enemy and he is us." -- Pogo
Greetings from coastal Washington state, the scenic US Pacific Northwest.