we currently have "low water in the sea", but (at least me) are crunching BIG stubs now (mostly >=200 GNodes the last day), so it should not empty that fast --> IMHO...
The proxies are running low again.
http://n0cgi.distributed.net/ogr-proxyinfo.html
I've noticed in the past hour or so that the free-dc proxy is not receiving any new stubs at all. Should be ok for a while....
Bok
we currently have "low water in the sea", but (at least me) are crunching BIG stubs now (mostly >=200 GNodes the last day), so it should not empty that fast --> IMHO...
the-mk
yes, this is 'today' on the proxy..
[root@quad dnetproxy]# ./csv5.pl 20081009
This might take a while... please be patient.
Overall OGRp2-25 PProxy stats based on 1 days:
Smallest: 0.45 Gnodes, submitted on 09/10/2008
Largest: 898.28 Gnodes, submitted on 09/10/2008
Average: 166.45 Gnodes
Total: 1,035,801.57 Gnodes
Rate: 1,035,801.57 Gnodes/day
Stubspace 7 stubs: 2,457
Stubspace 8 stubs: 3,765
Stubspace 9 stubs: 1
Total stubs crunched: 6,223 (0.00102% of project)
That average has been slowly dropping over the course of the day though, so perhaps smaller ones are on the way...hard to tell, it's like predicting the US stock market!
The keyservers are struggling from the looks of it. Only two are left with any workunits, and they seem to be emptying fast.
Let's hope the free-dc proxy gets us through
Keyserver's full again. Probably some older stubs that were recycled.
Engage!
At the end of the project it makes sense the keyservers are going to be a little overloaded. Hopefully as we get close to the end that there is not too much wasted work done with re-issued work that is almost completed somewhere else.
atm, looks like about half the servers have work. We're definitely in the endgame now.
WOOT!!! (Or, since I'm an Army brat, HOO AH!!!)
currently only 2 of 7 have a little bit of work to distribute...
it ain't over until the fat lady is singing
the-mk
All server's full again. Only 650.000 stubs to do, and the proxies still host their ususal 455.000 - what about all the stubs still in our clients an personal proxies. yoyo@home has a 50.000 stubs buffer, Bok's won't be to small either. It seems that there are more stubs now in circulation than are unresolved. I guess distributed.net is recycling faster than expected.
Engage!
I heard a rumour that the recycling of stubs was now down to every 24 hours. I'm probably not getting any of mine through that quick at all.
You could controversially say that we're now in a situation where stringing the project out would be better for us, because if we keep getting credit for already recycled stubs whilst MOOH runs his buffers to nothing, we could catch back up in the dying days. I'm not suggesting we do that though, it was just an observation.
Just changed my dnetc.ini file to this to see if I could snag the fresh wus from a keyserver, so don't panic if you don't see me on Bok's pproxy server as it is now listed at the bottom of the list for the time being.
[misc]
project-priority=OGR-P2=1,RC5-72=1
[networking]
autofindkeyserver=yes
nofallback=false
keyserver=*:2064;proxy1.dal.tx.us
keyserver=*:2064;proxy2.dal.tx.us
keyserver=*:2064;keymaster
keyserver=*:2064;proxy2.rwc.ca.us
keyserver=*:2064;proxy3.rwc.ca.us
keyserver=*:2064;proxy1.rwc.ca.us
keyserver=*:2064;proxy1.bris.qld.au
keyserver=*:2064;proxy1.kl.my
keyserver=*:2064;pproxy.free-dc.org
[OGR-P2]
core=-1
;fetch-time-threshold=24
fetch-workunit-threshold=-1
[rc5-72]
core=-1
;fetch-time-threshold=24
fetch-workunit-threshold=-1
I'm slowly reducing the cache on the free-dc proxy too.
all official proxies seem to be empty... how do we look like at Free-DC? are we drying up the pproxy?
the-mk
Yes,
it hasn't been able to get new stubs all day. Still has enough for quite some time yet though.
The average stubsize today seems to be about half of what it was yesterday as well, so if that trend extends out across everyone we could see a large chunk of the project finished today..
Bok
Well the dualie is now off-line so you'll see a massive drop in numbers from me, darn thing wouldn't remote control properly - i think this machine need a drop kick but I can't be bothered with it... and SoB is still not working so I've minimal numbers now...
Semi-retired from Free-DC...
I have some time to help.....
I need a new laptop,but who needs a laptop when you have a phone...
Now to remember my old computer specs..
The alleged 24-hour stub recycling seems to have done the trick this time round. It looks like we'll be ending the project almost at full throttle as opposed to what happened with OGR-24.
I realized that some of my online machines were not talking to the FreeDC proxy, so luckily I switched before my current batch ran out and they kept chugging away. Should be interesting to see what happens in another 20 minutes when the stats are processed if the project is over or not.
I had some issues with the free-dc proxy a day or so ago, but figured it was connected to the hdd issues the site was having (there was a thread about it somewhere).
I think we've got until the weekend 'til it's over. Looks like there is two day's worth of work going on current numbers.
And I want it to go on as long as possible, as I'm 524th as of yesterday's update. 24 places in two days. About 350 thousand Gnodes. I think I can, I think I can
The project has been running for 2997 days. It sure would be nice if we could end it on a very, very round number.
I still have plenty of work on my local pproxy and a small offline cache too.
99 bottles of beer on the wall 99 bottles of beer in one will fall We will have our forth place back
Are we there yet
I've still got a large buff_in.ogr file that I'm using some sneaky offline crunching for, plus whatever the free-dc proxy can provide. It would definitely be fantastic if we could finish on a number with lots of zeroes
% Complete: 99.97%
Daily Stubs Projected End Date
1 ...Day Average.. 132,962 ==17-Oct-2008
3 ...Day Average ..161,545 ==16-Oct-2008
7 ...Day Average.. 214,870 ==16-Oct-2008
14 .Day Average ..329,434 ==16-Oct-2008
30 .Day Average ..503,752== 15-Oct-2008
http://stats.distributed.net/project/ogr_status.php
BoK can you post here when the Free-DC PProxy is empty.
Just so i can switch remote boxes over.
absolutely...
I Gues the keyservers are out for good My client Buffers are bleeding out
We Only have 11 000.646 points to make back
I Think I Can I Think I Can I Think I Can
We Can Do It Go Team Go
Free-DC proxy is almost full again. They must have recycled again overnight.. Bok
Are the Key Servers up and running again ?
My Proxy server just refilled itself mmm I might need this again
If this keeps up I might make it to the top 10
Well if they were, they've been drained again. http://n0cgi.distributed.net/ogr-proxyinfo.html shows the keyservers status.
there are only 11 stubs left to crunch officially, so I doubt you just got "valid" work from the official keyserver. At least it would be very strange and useless (from an official point of view, not ours, of course).
Engage!
Don't you use the free-dc proxy?
Engage!
But doesn't your proxy server point to the free-dc one?.... It must do for you to show up here...
And the proxy is not quite dry yet...though it won't last that much longer, certainly through today and probably tomorrow.
Bok
I'm *sure* the free-dc proxy has more than yours....
Yes, LaurenU2 has a pproxy that holds ~20,000 to feed his hungry beasts.
Yes, his pproxy points directly to Free-DC
My PS3s have a cache of 1200 each.
It's not over until all the wus that have been handed out by the official proxy servers are completed.
Sure, the project might find that last wu and finish, but, points still count as the wus are finished out.
Some of you will say it's a waste of money/electricity and resources, and while I would agree with you normally, when the race is this close, I'll keep pharming this project until I run out of work, at which time my PS3s will switch to RC5.
If we can take back a spot, it will be worth the few more pennies in electricity to do so, and I'm not so fired up to be jumping on another project as others may be.
DPAD looks to be needing a little help, as we are loosing our huge lead in the dailies and the NPLB project looks like it could use a few cores of steady crunching.