There's debug code for determining the work cycle and who is due time, but need to know what client as it is for version 7 quite different to version 6 (7 runs work in large blocks before moving to next project).

Code to add to the cc_config.xml <log_flags> section:

<work_fetch_debug>1</work_fetch_debug>
<rr_simulation>1</rr_simulation>
<priority_debug>1</priority_debug>

--//--

P.S. You sure checked that you've not accidentally suspended work fetching from WCG?

Hello CPH,
There is no secret data in the Message tab. The first page of messages consists of data to post that describes your system. We often wish there were a way to get a lot more information posted to the Messages tab. There are some debug commands that post some more info.

Lawrence

Not sure you were joking or not, so here is: If your system is too busy with *non*-BOINC processes, BOINC will pause. The default setting is at 25%, so that is what is shown... a system busy with other things demanding a substantial amount of the CPU time, in bursts. The WCG default is actually 50% for new members, and 75% is good for most.

On forcing the hand to see if WCG work really will come: Suspend work fetches on your 4 other projects and sit back. Eventually there will be a gap in your cache minimum, which the client will want to back fill. Then it would try to get work from WCG or print log entries why it can't.

I've given you some log flags to add to the config to diagnose. Add to that, only working on clients 6.12 and earlier <debt_debug>1</debt_debug>. Which shows how much computing time each of your active projects is due. Maybe WCG is overworked... printing a negative number.

With 6.12.34 [Not client 7] you can also add the <zero_debt>1</zero_debt> to the cc_config.xml <options> section and read file config in. This resets all inter-project -debt balances to zero... i.e. clean slate, all getting 20%, for the present, as actual used time per project may vary due project task prioritizations.

Let us know

--//--

P.S., the Msg log told us that you've got BOINC installed at user level, not as service, and what your user account is (and that not in a million years I'd make same as a public member account). That, the device name and the external IP addresses are the only bits I'd advise to erase when posting message logs, to keep "secrete". When installed as service, BOINC runs in a limit rights account. Sample of what all my clients show:

6 30-6-2012 23:53:41 Running as a daemon (service)
7 30-6-2012 23:53:41 Data directory: F:\BOINC
8 30-6-2012 23:53:41 Running under account boinc_master

Because the boinc_master account can only run BOINC and access the BOINC data directory [sandboxed environment], no risk to share that.