In this thread...

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/forum_thread.php?id=1103

... I described my problems with CPDN after upgrading to the 4.13 BOINC core. I said at the time if SAH and LHC dry up I'd put 4.09 back on and reconnect to CPDN. They did and I have done this today.

My new CPDN unit has been running for 2½ hours so far, (other projects dettached or inhibited - thus it is at 100% at the moment). I will leave it running until tomorrow. Assuming it works, I will at a convenient point suspend processing, upgrade to 4.13 again and let it resume to see what happens.

I got the -5 error running on the 4.09 core as well. Then something weird happened, it looked like the entire BOINC system froze part way through downloading a new wu. I had to dettach and reattach.

Eventually, with a bit of manual intervention, I got a new wu. The new wu has submitted it's first trickle and has been running over 12 hours now.

It looks as though it could be a machine specific problem. My advice would be to keep regular backups and if it happens again it might be worth investigating the cause. Both OS and hardware problems generate this error message.

Same wu running, 43+ hours now.

I also had this problem. But since I set in preferences write to disk value to 50 climateprediction did not crash.