[RC5] client versions - AIX

Jason Hartzell jhartzell at arcataassoc.com
Wed Oct 2 09:37:56 EDT 2002


Well, I'm definately a stats junkie and I love the idea of the larger work
units. And I even still have some 133Mhz Pentiums working for d.net. I don't
need the machines fetching and flushing every 10 minutes, and the project is
long term. The stats will show up in the end. My daily stats will average
out and be fairly constant and I'll still get credit for all my work.

I don't see any problems.

-----Original Message-----
From: Timothy Marsh [mailto:Timothy.Marsh at usm.edu]
Sent: Wednesday, October 02, 2002 8:00 AM
To: rc5 at lists.distributed.net
Subject: Re: [RC5] client versions - AIX



----- Original Message -----
From: "Jack Beglinger" <jackb at guppy.us>
To: <rc5 at lists.distributed.net>
Sent: Tuesday, October 01, 2002 3:32 PM
Subject: RE: [RC5] client versions - AIX

> First off, a week is too long for any work unit.  It was also the most
current
> version available.   OGR has no appeal to me - for this very reason.

Since the same amount of credit is given no matter how long the unit is, it
seems to me that the only problem with having a longer unit is that it
delays the update of your stats until it is finished.  I can see how stats
junkies (I must admit I love stats too) would have a problem with this, but
I imagine most stats junkies also have some faster machines that don't take
more than a day to do an average OGR unit.  I don't think a week is that
long to wait.

> As I have mention prior too...  Give out bigger work units to personal
proxies.
> Let them handout smaller usable ones (2^28 if required, maybe memory only
> clients or older machines) to the local machines.  Then the personal
proxies
> can re-gather the information and hand back the big work unit to the
server in
> sky. If a part is missing - it can locally rehand it out to get it fixed
and the
> block big block returned.  This way, some of the large farms out there,
you
> could hand out 2^48 or larger and let them do the sub work.  This way your
> server are off loaded.
>
> Also if the personal proxy runs out of work units... Allow it to pick the
> random block to use.  This way it will "create" a block and hand it out to
the
> local crunchers.  In the end the personal proxy starts to appear to be the
> center of AMP machine, with one DNET client on each processor... no
> matter the type.
>
> At the same time let the personal proxies hand back pre processed stats.
> So checking a completed blocked from a personal proxy, also gives back
> tallies for the stats.

As stated in another reply to this email, this would require the pproxies to
do a lot more than proxy.  They would essentially be mini key-servers.
Also, Bruce mentioned earlier that the stats database has to keep at least
one bit for every one of the smallest units.  I believe that means that if
we have small units (say 2^28), then even if someone processes a larger
block, it will be stored in the database as many smaller blocks.  Even if
somehow the pproxy could assemble the stats, it would have to report the
work done by different people (if different people use the same pproxy) as
fractions of a larger block, which I don't think would work.

Besides, I don't believe that most of dnet participants use a pproxy any
way.

--
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 mailing list