Commit Graph

12 Commits

Author SHA1 Message Date
David Anderson 8c591e31df - client: first whack at new work-fetch logic. Very preliminary.
svn path=/trunk/boinc/; revision=16754
2008-12-31 23:07:59 +00:00
David Anderson 2dc7056ee0 - client: code shuffling
- scheduler: fix typo in msg

svn path=/trunk/boinc/; revision=16750
2008-12-30 19:01:25 +00:00
David Anderson b24d947aaa - client: work fetch: skip project if zero CPU shortfall
and currently have coproc jobs.

svn path=/trunk/boinc/; revision=16746
2008-12-27 03:43:13 +00:00
David Anderson ea0146d154 - client: fix calculation of CPU shortfall;
don't fetch work from projects with zero CPU shortfall

svn path=/trunk/boinc/; revision=16613
2008-12-03 23:30:54 +00:00
David Anderson 79fb6e969e - Remove the notion of "CPU efficiency" from both client and server.
This wasn't being measured correctly for coproc/multithread apps,
    and its effect is now subsumed in DCF.

svn path=/trunk/boinc/; revision=16610
2008-12-03 19:50:06 +00:00
David Anderson 89548f04da - client: compute duration_correction_factor based on elapsed time, not CPU time
(otherwise it doesn't work for coproc or multi-proc apps)
    - client: in estimate of job completion time,
        weight the estimate based on fraction done more heavily
        (quadratic rather than linear)

svn path=/trunk/boinc/; revision=16603
2008-12-02 22:19:39 +00:00
David Anderson 122f61996a - client: estimate time to completion based on elapsed time, not CPU time
(improves estimates for coproc apps)

svn path=/trunk/boinc/; revision=16600
2008-12-02 17:48:29 +00:00
David Anderson 78199c2f5e compile fixes
svn path=/trunk/boinc/; revision=16599
2008-12-02 16:48:25 +00:00
David Anderson 84f1193a9d - client: use FLOPs, rather than CPU time,
as the basis for estimating job completion times.
    This should improve estimates for GPU apps,
    and prevent the DCF from getting messed up.

svn path=/trunk/boinc/; revision=16598
2008-12-02 03:58:32 +00:00
David Anderson eae866f743 - client: for CPU throttling, don't suspend apps that
are non-CPU-intensive or that use < 1 CPU (e.g., CUDA)
	- client: get rid of spurious "internal error,
		expected process to be executing" msg
	- diag: don't check heap on every alloc
	- fix a few compile warnings

svn path=/trunk/boinc/; revision=16323
2008-10-27 20:17:22 +00:00
David Anderson e24e551bd8 - client: clarify and fix the semantics of "next RPC time".
Here's are the new semantics: a scheduler reply can include
        <next_rpc_delay>
            Make another RPC ASAP after this amount of time elapses.
            This is specified by the <next_rpc_delay> element in config.xml.
        <request_delay>
            Don't make another RPC until this amount of time elapses.
            This is sent automatically (and sometimes with large delays)
            by various parts of the scheduler.
        next_rpc_delay now "overrides" request_delay in the sense that
        request_delay is ignored if it's greater than next_rpc_delay.

        In addition: the client maintains a min_rpc_time which is set based
        on request_delay and also by various exponential backoff schemes.
        new_rpc_delay now overrides this as well, in the same sense.

svn path=/trunk/boinc/; revision=16206
2008-10-14 21:16:04 +00:00
David Anderson 98cfb8d3b0 - rename .C files to .cpp so that Doxygen will work
svn path=/trunk/boinc/; revision=16069
2008-09-26 18:20:24 +00:00