LHC@home

Just remember that your reliability score will go down as you increase the buffer. The reliability score should improve over time as you decrease buffer.
Low reliability means your computer will be sent less work in general and less mission critical work (soon to expire or if a third result is needed).

Please don't hesitate to correct me if I am misinterpreting these values or add where you can provide the project specific parameters and thresholds.

Just remember that your reliability score will go down as you increase the buffer.

hm, is this really the case? It's the first time I hear about this kind of provision ...

http://boinc.berkeley.edu/trac/wiki/ProjectOptions#Acceleratingretries

It is part of the documentation. I assume it's still in use, because, in my purely anecdotal experience, I did receive high priority WUs when my system was crunching reliably. Since November I have seen fewer such WUs. Of course it all depends on the parameters specified by LHC and how they setup the project. Maybe I'm simply experiencing a queuing issue atm.

In times when WUs are plenty I don't expect issues receiving work for any host.

Thanks to the IT infrastructure guys, the situation is getting better - please see:
https://lhcathome.cern.ch/lhcathome/forum_thread.php?id=4567#33713