[RC5] Packet Lifetime

Bruce Wilson bwilson at distributed.net
Thu Aug 9 22:23:23 EDT 2001

A fair question, and a good one.  The administrative work needed to
recycle OGR-24 is quite substantial, and only certain people can do it.

The work units which have been returned multiple times can only be
considered "complete" if they are submitted by different participants,
and those participants return exactly the same results.  In this way, we
can be sure the work is not damaged in transit, and that calculation
errors (known or especially unknown) will not spoil the scientific
significance of the project.

The keymaster is not set up to maintain the participant information with
each work unit returned, and we would not wish it to keep track of such
details.  Currently, we are forced to reload all the OGR logs since the
inception of OGR-24 (or -25, or...) into a large SQL database to perform
several large queries on the data.  The storage requirements are huge,
and processing time is also quite intensive.

With tally as fragile as it is, and blower still in flux (we're waiting
on Sybase for an answer), we haven't had a good option for performing
all the processing.  blower is functional enough that we could attempt
the processing, but then personal schedules have also become a limiting
factor.  For much of the OGR project, Daa has been the only person with
the skills, ability and availability to perform all this processing, but
his availability has been severely limited as of late.  He recently
explained the process to me and gave me his scripts, but I have lacked
(and still lack) the "availability" part of the equation.

Obviously, we understand the importance of getting OGR-24 back on track,
of keeping -25 on track, and of getting -26 started.  The facts of my
situation are that it is unrealistic for me to even attempt the recycle
until sometime in mid to late September, and that time could be eaten up
before I get there.

We are also looking for ways to streamline or automate this, but OGR-24
is so overdue for a recycle that it would be foolish of us to focus on
automation at this time.  If I am the next one to manually recycle OGR,
I will be looking at options for full automation.

I remain fully committed to making this happen to the best of my
ability.  I know the others feel as I do, but with the same constraints
on their time.  I won't apologize for giving my family and my paying job
the priorities they deserve, and I won't criticize Daa for doing the

We all want distributed.net to succeed, but the only way we can
realistically continue to be involved at all is to keep paying the
bills, satisfying our paying clients, spending time with our
mostly-patient families, and trying to find a healthy balance in our
lives.  Getting fired, going broke, alienating family or losing our
minds won't help distributed.net.

If anyone reading this list has experience developing for Sybase SQL
Server, with some Unix shell script ability, and would be willing to
become a full member of DCTI, please e-mail me directly
(bwilson at distributed.net) so we can get you involved.  Experience with
PHP, VLDBs, Perl, and ssh are all helpful, but not required.  Enthusiasm
for the project is a must.  We are continuing to search for other
volunteers who can help spread out the workload, especially in stats,
core development and general client development, but all comers are
welcome.  I put the call out pretty frequently on this list, but it is
rare that I get even one response.

(Anybody who happens to work for Sybase and could maybe nudge someone to
get us an answer would also be greatly appreciated.)
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 Tom Ribbens
| Sent: Thursday, August 09, 2001 17:28
| To: rc5 at lists.distributed.net
| Subject: Re: [RC5] Packet Lifetime
| Mike Henry wrote:
| <snip>
| > 
| > OGR-24 started 392 days ago, but it's just dribbling along. 
| It's sad and
| > depressing to watch, and I don't think it's good for 
| distributed.net.
| > Only 8 people submitted work yesterday according to the stats page.
| > 
| > We are being held to ransom by a few people with slow 
| machines (I blame
| > the client for having a default which causes you to download FAR too
| > much OGR work) - it'll be months before they submit all the 
| blocks they
| > are holding.
| > 
| > Please, please, PLEASE can't we just say enough is enough, 
| recycle the
| > remaining OGR-24 blocks, finish it off, and give it a 
| dignified burial?
| Here's a question that i never have seen asked, but might be 
| intresting 
| to know about the recycling:
| How much work is it to recycle those OGR-24 stubs? I mean both 
| administrative (as in live humans jobs) and computionally 
| (generation of 
| the list of stubs to hand out again).
| > 
| > Cheers
| > Mike
| Greetz
| Tom
| --
| 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