24 hours or some lesser value is used also to back-off badly behaving clients who're down to their quota, which resets at midnight. The WCG version is kept as close as possibly true to the Berkeley version, so as not to confuse members, so would not favor different scheduling behavior, WCG being able to tell clients how long to back off... like the 11 seconds.

--//--

We are starting another stress test of cep2 beta. We have changed a couple of things thought.

1) They will run for 2 hours max (we aren't using the results for these, just stressing the download servers). The estimate time has been reduced as well.

2) We are limiting the max number of connections to the download servers. The last beta reached a max overall bandwidth of 120Mbps but it was divided over so many connections that it was very slow for all. We will see if this improves the overall throughput.

That has given me something to seriously consider and probably explains why I am reading about some people managing to grab many betas while I am grabbing only a few. It seemed to be different during the last beta flood but when I think about it, I was using my beta-profile consistently when there were many betas to be had.

I think it would be better for me to stop casting my 'beta-profile net' into the waters until I see that there is another flood.

Thanks for your advice.

EDIT:
...and behold, he replaced his beta-profile net and almost immediately four juicy betas jumped up onto his boats

Thanks... got the max, of both :D

Now it's shuteye time.

--//--

Looks like a flood to me, picked up 83 2h tasks... :)