I have two projects with 100/15 resource share. I've set BOINC not to keep buffer of tasks for projects.
According to event log, project with 100 resource share is completely ignored by BOINC. This is what it does on its internal timer, and the same thing happens when I try to Update the project.

30/03/2013 02:04:20 | LHC@home 1.0 | Sending scheduler request: Requested by project.
30/03/2013 02:04:20 | LHC@home 1.0 | Not reporting or requesting tasks
30/03/2013 02:04:22 | LHC@home 1.0 | Scheduler request completed

How do I force it to at least attempt to fetch tasks for the project? It completely ignores me.

a) Which BOINC?
b) How long has it been doing that?
c) What's the scheduling priority value for LHC and the same value for the unknown other project?
d) How long is the deadline on both projects?
e) How many days worth of work do you ask for?
f) Where's the log with the <cpu_sched_debug/> and <work_fetch_debug/> flags?
g) Does LHC even have work at this time?
h) How much work do you have in cache for either project?
i) Want more questions? :-)

---