Now, the rest of the story.... My helpful colleagues (thanks, Dave and Wolfie!) found a sequence block with a couple of divided-by-zero calculations that get sent out as "infinity" in the real outputs. (The zero is the feed setpoint, imagine that, during a process shutdown!) Once we fixed that error, 09CP02 uploaded flawlessly, albeit still taking 40 minutes to do so. Meanwhile, I found some undefined blocks trying to read Modbus inputs via an FBM232 in the other CP, 10CP02. As they were built for testing, I nuked them away, and the problem ... kinda went away. The first upload after I deleted those blocks failed, but I tried it again and it was successful. No, I didn't have any good explanation for that either. Thanks for all the suggestions. And thanks, Alex, for sending me the procedure to remedy database mismatches. I didn't need it this time, but it can be handy in the future. (Oh wait, I wrote a procedure to solve that problem once myself. Need to find that one.) Regards, Duc _______________________________________________________________________ This mailing list is neither sponsored nor endorsed by Invensys Process Systems (formerly The Foxboro Company). Use the info you obtain here at your own risks. Read http://www.thecassandraproject.org/disclaimer.html foxboro mailing list: //www.freelists.org/list/foxboro to subscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=join to unsubscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave