I wonder why seti@home keeps pushing work units to my main system? Second system (2nd workhorse) wasn't getting anymore units at all. Thought i'd wind down my primary system, but it keeps getting units pushed to it. Very weird.
Is it common for the projects to favor a system over another?
I don't know about favoring one system over another in general, but it is possible to set a project up so it only sends work to faster hosts. SETI might be starting that so things don't linger too long past their March 31st deadline...I doubt they want slower hosts to linger for 2 months grinding away on a large cache of WU.
Interesting. The secondary is hardly slow, it's a newer I-5 versus the older dell workstation with 8 cores and a gpu. The other dell has 4 cores and a nvidia versus amd gpu. Just got another 3-4 units (and many before that even). Real strange.
Oh well, einstein benefits off the 3020 not getting units versus the T7500 getting seti@home still
Hello HAL here, another refugee from SETI. I'm devoting all the resources of the Linux machine I was using for that project (Ryzen 7 1700) to this project. Congratulations!
Looks like you have a word association thread too, that'll keep be busy here LOL. Cool
Processing work units with "outdated" (according to Microsoft) Ryzen 7 1700
There's something they did that seems to favor slower machines over faster ones when work is scarce. I see that here. The 359k rac machine takes 3-4 days now to get a full cache after the outage, and is first to run out. The 250k rac machine the same but less so, and the 100k and 20k machines fill up in no time. It's more than luck of the draw, for sure.
There's something they did that seems to favor slower machines over faster ones when work is scarce. I see that here. The 359k rac machine takes 3-4 days now to get a full cache after the outage, and is first to run out. The 250k rac machine the same but less so, and the 100k and 20k machines fill up in no time. It's more than luck of the draw, for sure.
Seti was always the first to test new Server side settings ie sending resends to faster computers so the workunit is done with quicker, but not all of them worked as expected for various reasons. You could be seeing something like that going on.
Indeed we will.
)
Indeed we will.
On a happier
)
On a happier note: boinc.berkeley.edu refused to connect.
Often this has signaled the outrage is getting ready to end.
SETI refugee landing in
)
SETI refugee landing in 3...2...1...
I'm guessing we will see a
)
I'm guessing we will see a lot more visitors here now and not just on Tuesday's. Welcome to the insanity.
I wonder why seti@home keeps
)
I wonder why seti@home keeps pushing work units to my main system? Second system (2nd workhorse) wasn't getting anymore units at all. Thought i'd wind down my primary system, but it keeps getting units pushed to it. Very weird.
Is it common for the projects to favor a system over another?
I don't know about favoring
)
I don't know about favoring one system over another in general, but it is possible to set a project up so it only sends work to faster hosts. SETI might be starting that so things don't linger too long past their March 31st deadline...I doubt they want slower hosts to linger for 2 months grinding away on a large cache of WU.
That's how I'd do it, but it's just conjecture.
Interesting. The secondary is
)
Interesting. The secondary is hardly slow, it's a newer I-5 versus the older dell workstation with 8 cores and a gpu. The other dell has 4 cores and a nvidia versus amd gpu. Just got another 3-4 units (and many before that even). Real strange.
Oh well, einstein benefits off the 3020 not getting units versus the T7500 getting seti@home still
Hello HAL here, another
)
Hello HAL here, another refugee from SETI. I'm devoting all the resources of the Linux machine I was using for that project (Ryzen 7 1700) to this project. Congratulations!
Looks like you have a word association thread too, that'll keep be busy here LOL. Cool
Processing work units with "outdated" (according to Microsoft) Ryzen 7 1700
There's something they did
)
There's something they did that seems to favor slower machines over faster ones when work is scarce. I see that here. The 359k rac machine takes 3-4 days now to get a full cache after the outage, and is first to run out. The 250k rac machine the same but less so, and the 100k and 20k machines fill up in no time. It's more than luck of the draw, for sure.
Jimbocous wrote:There's
)
Seti was always the first to test new Server side settings ie sending resends to faster computers so the workunit is done with quicker, but not all of them worked as expected for various reasons. You could be seeing something like that going on.