garoob,
The fact that you can see only 5 days of work is not because you have a dual-core machine. If you had a single core you would see half of them! The reason is due to the way Boinc integrates your connection rate and the deadlines of the jobs for evaluating how many WUs it can stack without missing deadlines. To make it a little more complex (you are allowed to read it "unpredictable") this algorithm is not the same in all versions of Boinc. The practical bottom line is that you will hardly ever get more than 5 days of work in Boinc's queue for WCG projects. So if you are already getting that much you are already at your best!

Cheers. Jean.

It's undocumented and only found out thru asking. The cc_config.xml is still very much valid in itself. Use it all the time and it should be there to control the level of messaging detail you get. Maybe you have it malformed somehow.

Here's a good sample with the iimmobilized line struck-out.


With the Additional Buffer of 5.10.x you can get real maximum 10 days worth of work in the buffer. Just set connect to 1.0 days (default WCG 0.3, default BOINC 0.1) and additional buffer from 1.00 to 10.00 days (WCG has no default yet, BOINC default is 0.25 days). What you will find if you stretch the value toward the max, your client will run continuously in "running, high priority" (5.10.x series only message). Not recommended to do unless you really need it. The lengths of work are highly variable. Just uploaded FAAH jobs from 1 client that varied from 4:25 to 7:20 hours. BOINC continuously adjust it's estimations on last work completed for a project. Suggest you visit the various wikis out there for elaborations. They are not up to snuff with latest scheduler logic, so just take it as a broad line informational.