This challenge is to see how well EDGeS can hold up to increased resources due to challenge stress. Please join in and report your experiences both with server response and other project issues so that we can better determine whether it is a good fit for the DC-Vault.

The ever present 'New User' from team 'New User' has entered the premises...

Faster Scotty: Results ready to send 37
Results in progress 3,676
Workunits waiting for validation 1
Workunits waiting for assimilation 23
Workunits waiting for deletion 0
Results waiting for deletion 0
Transitioner backlog (hours) 0

yip, I keep getting project has no tasks available and then my backup projects kick in. Not getting a constant source of units.

This is good to know. Lack of work to feed everyone is certainly a concern. Has anyone found any problems with the work units they have received? In the past the Zeta work units didn't always suspend as directed by the client...

I don't think the units checkpoint as they seem to start from scratch if Boinc is shut down. Either that or the wrapper starts the time from 0 on restart.

The zeta WUs do not suspend. They stay active. This is on an Intel CPU, WIN 8.1, BOINC 7.4.22

Seems you can't have more WUs than threads. I can confirm they restart with boinc restart. The Riemann Zeta Research Project has three applications; I presume when I get "no new tasks" and the server still has WUs to send it's the Linux versions of the project waiting to be sent as its probably out of the more popular windows WUs.

The project lets itself be run under a low priority. Even though I've set resources to 10,000% it will be 'pushed away' by SRBase set at 1,000%.

Results ready to send 0
Results in progress 3,248
Workunits waiting for validation 2
Workunits waiting for assimilation 2
Workunits waiting for deletion 0
Results waiting for deletion 0
Transitioner backlog (hours) 0

At home.edges-grid.eu/home/performance/wu_unsent14day.jpg you can see a graphical representation of the number of unsent WUs (highly optimistic)

I've seen the same Dirk. eDGES at 10000, Finds dDM and SRBase at 50 all get downloaded and run. Basically setting priority with this project doesn't work. All my backup projects run in preference regardless of what I set.

Great info guys. Keep up the good work. All of this helps us determine its viability and will be relayed. I plan on using these challenges in the future to help us assess and stress test. Only so much a hand full of people can do on their own while testing.

Having to constantly update the project myself to try and get very limited amounts of work. Basically I have a 8 core xeon and 8 core i7 sitting idle. Project works as a backup to other projects, not much use as a competition or main project.

I posted some of the issues in their forum to give them an opportunity to address them. I doubt they are ironed out by the time the challenge is over, but at least they are aware of what we are seeing.

On my account it states that my max work units is 500 per core but the project will allow no more than 2 units per usable core to be downloaded. So, on say my i7, the max work I can download is 30 mins worth of work and that is provided the project has any units to send.

Results ready to send 575
Results in progress 3,333
Workunits waiting for validation 0
Workunits waiting for assimilation 1
Workunits waiting for deletion 0
Results waiting for deletion 0
Transitioner backlog (hours) 0

Project server seems to like SETI.USA, they get almost all the work

Anyone notice how many invalid WUs there are? I have a ton of them on all systems.

I never noticed I had so many either. They are all marked invalid with a successful outcome and valid exit. I can see no difference between the units marked valid and the units marked invalid ? Is the project simply accepting 2 of the 3 produced units valid and marking the 3rd unit invalid ?
This is not good, as everything I get is running successfully before turning out to be a waste of time. Adds to all the wasted time I am getting when my systems are sitting idle waiting on work.

scole of TSBT, I'm not noticing a high rate of invalids. What do you consider a "ton"?

alez [ TSBT's Pirate ] I was just mentioning that observation to my team last night.

actually more worrying, from checking my systems if a unit gets sent to a win box and a linux box whichever operating system gets sent the 3rd will determine which 2 units validate. linux always validates linux and win marked invalid. Same win always validates win and linux marked invalid. This is a huge issue with the project and a massive waste of our resources.

as SZTAKI Desktop Grid does the same work, think I will be finished with this project after the challenge. Looking through the forum, the admins are never around and the validate issue was reported 10 months ago without reply. So far I think this project has too many issues to be regarded as a candidate for your vault.

About 80 invalid, 1700 valid on the 6th. 50 invalid, 800 valid so far today, the 7th. That's around a 5% invalid rate.

Thanks for the info guys. This is very helpful.