OC,

Looks like you might be right about SETI - Germany switching resources to CEP. They increased their daily production by 62K points on CEP yesterday and decreased daily productivity on the other 3 WCG projects.

We are showing a 501K points increase in daily production for yesterday while SETI-Germany is showing a further drop of 562K in production. I expect that all the recently added crunching resources to our XtremeSystems Team will kick in big time tomorrow and the day after. Hopefully we can erase the deficit and build enough of a point lead to counter a massive SETI Germany diversion to CEP WUs and their delay reporting. A 3.5 millions point lead or greater in our favor, at the end of this contest, appears to be a desirable safety margin.

Have you been able to find out exactly when this contest will end is it the 25th @ 23:59 hrs. or the 27th @ 23:59 hrs.?

Hi Xtremers....

For your information realting to the stats.
The challenge includes all the points validated during the event time.
So the points shown in our challenge are a result of a special hourly export, that is done by the WCG-techs for this event only.
The export includes all the points a team has get granted in relation to the points of the last midnight update when a team joins the challenge.
So our stats follows the same way, like the challenges are calculated on the grid.
Because of the possibility of changing a team, the special export is not able to cover this at all. After a following midnight update (mostly at 2.35 in our stats) the output of the teams will be corrected automatically by the next hourly export. That is the reason, why our stats do not have the same daylie output value on the grid challenge site, because this is done on 2.35 (the next day). But the totals are safe! :D

At the end, with a final script, we are fetching the data of the last midnight export and bring it in relation to the first available midnight export of every team. With that data, we are sure, that all the points generated during the event, will be counted on our stats and will be shown the same points.

Hope this helps!

And of course: Thanks for participating in this event!
It seems, that it will be a big final!! :D

Ok, the XS forum is having a little "issue" at the moment so I'll post this here.

In response to OC's post regarding a diskless machine running of a USB stick with persistence:

Ok, this might prove useful to people so here's what I just did that seems to be working.

I just got an install of Ubuntu Server (not desktop, just commend line) running with full persistence on a USB stick.
How?

Create the install media the same as usual. CD/DVD/USB, whatever suits you. I used a 4GB USB stick.
Plug both the install stick and another stick, probably a 32GB one (might work with a 16, I don't know) into the target system
Boot from the install media and go through the process as normal.
The installer will ask you to unmount local storage. Say yes (but don't unplug anything)
Direct the installer to use the USB stick as the "hard drive" it's installing to.
Use the "Use entire disk - Guided Install" option.
Let it partition the drive and install etc as normal.
When the option comes up make sure you select to install the SSHServer package (up and down cursor keys, space bar to select the option, enter to commit)
Remove the install media when it's done but leave the 32GB stick in place and reboot.
Update and install boinc as normal.

I did this with an AMD A6-3600 APU, generic HP motherboard, 8GB RAM combo. It returns less points than a Q6600 but also uses less power so is very quiet.

That's a good question. I wouldn't want to use it while running CEP2, for example. The high write volume would chew the stick up in no time and the low IO speed (of a USB 2.0 device) would kill machine efficiency horribly. A USB 3.0 system might not have the performance problem.

I posted what I did as it's so similar to installing to a regular HDD it surprised me. No need for special config files or extra partitions, just install and run like normal. I only tried it to see what errors I would get and it seems to be working just fine.

I shall continue to try to get to know Fatdog 64 and its idiosyncrasies as it seems that the io issue is resolved by default. It runs entirely in ram, has a very small footprint and seemingly has an option on shutdown of moving everything to a savefile. Works great as a throw away/temp machine as is.