[RC5] possible third project...
bwilson at distributed.net
Sun Jul 29 15:47:51 EDT 2001
The client buffer is pseudo-FIFO. When the client finishes a work unit,
it selects the next work unit based on the work unit identification
number, regardless of the order it was received. In selecting the work
unit with the highest (or lowest, I forget which) work unit number, it
attempts to complete the oldest work first.
I know this can have an impact on a recycle, because this was part of
the problem we encountered with CSC. When we started over at the first
block, the client "preferred" the earlier blocks over the later blocks,
and it took a long time for the last percentages to be returned. I
don't know if this has been resolved in the OGR projects, or if it has
the impact you describe. I would hope that OGR-24 would be treated as
"older" than OGR-25 by dnetc, but I'm not the one to ask.
Bruce Wilson, Manager, FERS Business Services
bwilson at fers.com, 312.245.1750, http://www.fers.com/
PGP KeyID: 5430B995, http://www.toomuchblue.com/
Domain names are like women (or men): The ones you want are
already taken, but you can always get one from a strange country.
| -----Original Message-----
| From: owner-rc5 at lists.distributed.net
| [mailto:owner-rc5 at lists.distributed.net] On Behalf Of Michael
| Sent: Sunday, July 29, 2001 12:31
| To: rc5 at lists.distributed.net
| Subject: Re: [RC5] possible third project...
| --- Tom Ribbens <webmaster at multi-air.com> wrote:
| > Mike Henry wrote:
| > > On Fri, 27 Jul 2001 21:21:12 -0400, Greg Wooledge wrote:
| > >
| > ...
| > >
| > > Or we could work on OGR-24 for a few days and finish it next week
| > (HINT
| > > HINT!)
| > and _then_ start a new contest...
| The only way we'd finish it next week is by handing out one
| each of the remaining OGR-24 stubs to each connecting client,
| and hoping that it winds up on top of the input buffer and
| that one of the clients in question is not a 386.
| That brings up another question I had -- is the OGR buffer
| FIFO or LIFO? I've gotten an unscientific impression that
| it's LIFO, meaning that buffer update option 2, ensure
| completed work is kept flushed -- which also fetches another
| stub to fill out the buffer when the finished one is flushed
| -- keeps the remaining stubs buried.
| Or is it just my imagination?
| -Michael Pelletier.
| Do You Yahoo!?
| Make international calls for as low as $.04/minute with
| Yahoo! Messenger http://phonecard.yahoo.com/
| To unsubscribe, send 'unsubscribe rc5' to
| majordomo at lists.distributed.net rc5-digest subscribers
| replace rc5 with rc5-digest
To unsubscribe, send 'unsubscribe rc5' to majordomo at lists.distributed.net
rc5-digest subscribers replace rc5 with rc5-digest
More information about the rc5