[. Would you allow me to join your party over there? It wouldn't amount to very much, but it may as well go to "help" a team, and I consider you and CDNgeezer as friends, so ...
Michael R.
>It would be great to have you over there, Michael.....The Admins try very hard to keep everyone 'in the loop' and it may be a pleasant surprise in comparison. Every contribution big or small is really appreciated as you will see.
PS...I am very honoured that you include me as a friend....Cheers, Rog.
CDNgeezer & tweakster:
I am in.
How do I set the relative resource share between the two projects? I have never run anything other than E@H.
>Hi Mark....it would be great to have your help with Rosetta and with a second project you won't have to monitor E@H so closely for 'the curse'. Here is how it works: your account page on either project has a heading called 'Einstein@Home Preferences' or 'Rosetta@Home Preferences'. When opened, there is a statement called 'Resource Share' where you set up whatever values you need so BOINC can create a resource ratio between projects.....ie if E@H ResShare is set at 100 and you set Rosetta for 100 ResShare also, BOINC will give you a 50%-50% operating time split between them. If one project is set at 80, the other at 20 you will get a 80%-20% split, and so on....You can change the values when the 'edit' button is activated and then saved when you hit the 'update preferences ' button. You will then get a message that informs you when the changes will take place. Hope this helps...Cheers, Rog.
Mike; The purpose of the "Curse of 32" team at Rosetta is to let the Einstein Admins take a look at the crunch power they are wasting.....
You are a true hard core cruncher!!! I mean that nicely, Tweakster. :-)
My guess is that the concern is the stability of their overall system, of which the server is central, in order to ensure or defend the quality of the science result.
It's an interesting scenario actually. Over at Stardust they have significantly delayed the start of the project due to overwhelming demand. ( It starts this week by the way ). The registrations of interest they recieved far exceeded the capacity of their original IT design, and thus required reconfiguring. But at least they had the benefit of some advance warning! Back here at E@H they've been totally blindsided by Akos Fekete's ( all hail !! ) brilliant work. So very likely there's a tension on the one hand between wanting to satisfy and keep users of all varieties, but without risking the integrity of the project overall. If you are implying that the project administrators are indifferent to user concerns generally, I think that is a bit unfair. Witness their hiring of Akos for instance, or that they have doubled the WU quota twice already to get to 32 per day ( since I joined in Dec '05 ). Maybe they just can't oblige for practical reasons, pure and simple.
Please do keep any critique's coming though, not that you need any encouraging.... :-)
I personally am amazed by the refreshing frankness of discussions here, and I would welcome all opinions on E@H ( I only moderate the mode of interaction ..... )
Cheers, Mike.
(edit) In an indirect way, maybe Akos has helped other projects too!!! While enabling a fourfold-plus increase in throughput for E@H, that also leaves time over for others like Rosetta etc... in the manner that Michael Roycraft points out. The science value is easily equal or better even. While it's terrific that you are/were so devoted to E@H crunching, don't forget that BOINC is designed to spread the work around between whatever projects that a user feels is worthy. I think that's probably a key point - to do what you feel is best - in my view as a volunteer you ought to satisfy your own script here and don't needlessly feel obliged.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
Mike:
I normally don't feel any need to respond your excellent posts. But in this case I disagree on two points.
First it isn't just 32/day/cpu from my observation there is also a 4 CPU cap. Increasing that would be unlikely to destabilize the server.
Second there has been no response from the devs. Bring Akos in to help with the optimization is great and I expect good things to come from it. However the protracted official silence looks like they just don't care enough to respond.
Actually I suspect the devs are somewhat relieved. If things continue as they are they could run out of work in a couple of weeks. 7*1,615,524 /685,305~16.5 days
Mike:
I normally don't feel any need to respond your excellent posts. But in this case I disagree on two points.
First it isn't just 32/day/cpu from my observation there is also a 4 CPU cap. Increasing that would be unlikely to destabilize the server.
Second there has been no response from the devs. Bring Akos in to help with the optimization is great and I expect good things to come from it. However the protracted official silence looks like they just don't care enough to respond.
Actually I suspect the devs are somewhat relieved. If things continue as they are they could run out of work in a couple of weeks. 7*1,615,524 /685,305~16.5 days
Your are quite right, I don't actually know, and yes some official word would be good. Please don't get any feeling that I automatically defend the project, simply because I've been 'awarded' moderator 'status' or whatever, I'm really just a hopeless optimist who tries to look on the bright side ..... :-)
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
LOL, you are a mod; congrats. I had not noticed until you mentioned it.
Ha! :-)
It's just that since doing this moderating stuff, some posters have thought that I represent the project in some official sense, so I try not look like I do - so I don't misrepresent anything for any party! I am an absolute volunteer ( another continent ) with no special inside link/knowledge to the project....
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
So what does it take to disperse workload across several machines while still maintaining a common and coherent database? So as to not have all the breakable eggs in one breakable basket....
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
RE: [. Would you allow me
)
>It would be great to have you over there, Michael.....The Admins try very hard to keep everyone 'in the loop' and it may be a pleasant surprise in comparison. Every contribution big or small is really appreciated as you will see.
PS...I am very honoured that you include me as a friend....Cheers, Rog.
CDNgeezer & tweakster: I am
)
CDNgeezer & tweakster:
I am in.
How do I set the relative resource share between the two projects? I have never run anything other than E@H.
RE: CDNgeezer &
)
>Hi Mark....it would be great to have your help with Rosetta and with a second project you won't have to monitor E@H so closely for 'the curse'. Here is how it works: your account page on either project has a heading called 'Einstein@Home Preferences' or 'Rosetta@Home Preferences'. When opened, there is a statement called 'Resource Share' where you set up whatever values you need so BOINC can create a resource ratio between projects.....ie if E@H ResShare is set at 100 and you set Rosetta for 100 ResShare also, BOINC will give you a 50%-50% operating time split between them. If one project is set at 80, the other at 20 you will get a 80%-20% split, and so on....You can change the values when the 'edit' button is activated and then saved when you hit the 'update preferences ' button. You will then get a message that informs you when the changes will take place. Hope this helps...Cheers, Rog.
got it thx
)
got it thx
RE: Mike; The purpose of
)
You are a true hard core cruncher!!! I mean that nicely, Tweakster. :-)
My guess is that the concern is the stability of their overall system, of which the server is central, in order to ensure or defend the quality of the science result.
It's an interesting scenario actually. Over at Stardust they have significantly delayed the start of the project due to overwhelming demand. ( It starts this week by the way ). The registrations of interest they recieved far exceeded the capacity of their original IT design, and thus required reconfiguring. But at least they had the benefit of some advance warning! Back here at E@H they've been totally blindsided by Akos Fekete's ( all hail !! ) brilliant work. So very likely there's a tension on the one hand between wanting to satisfy and keep users of all varieties, but without risking the integrity of the project overall.
If you are implying that the project administrators are indifferent to user concerns generally, I think that is a bit unfair. Witness their hiring of Akos for instance, or that they have doubled the WU quota twice already to get to 32 per day ( since I joined in Dec '05 ). Maybe they just can't oblige for practical reasons, pure and simple.
Please do keep any critique's coming though, not that you need any encouraging.... :-)
I personally am amazed by the refreshing frankness of discussions here, and I would welcome all opinions on E@H ( I only moderate the mode of interaction ..... )
Cheers, Mike.
(edit) In an indirect way, maybe Akos has helped other projects too!!! While enabling a fourfold-plus increase in throughput for E@H, that also leaves time over for others like Rosetta etc... in the manner that Michael Roycraft points out. The science value is easily equal or better even. While it's terrific that you are/were so devoted to E@H crunching, don't forget that BOINC is designed to spread the work around between whatever projects that a user feels is worthy. I think that's probably a key point - to do what you feel is best - in my view as a volunteer you ought to satisfy your own script here and don't needlessly feel obliged.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
Mike: I normally don't feel
)
Mike:
I normally don't feel any need to respond your excellent posts. But in this case I disagree on two points.
First it isn't just 32/day/cpu from my observation there is also a 4 CPU cap. Increasing that would be unlikely to destabilize the server.
Second there has been no response from the devs. Bring Akos in to help with the optimization is great and I expect good things to come from it. However the protracted official silence looks like they just don't care enough to respond.
Actually I suspect the devs are somewhat relieved. If things continue as they are they could run out of work in a couple of weeks. 7*1,615,524 /685,305~16.5 days
RE: Mike: I normally don't
)
Your are quite right, I don't actually know, and yes some official word would be good. Please don't get any feeling that I automatically defend the project, simply because I've been 'awarded' moderator 'status' or whatever, I'm really just a hopeless optimist who tries to look on the bright side ..... :-)
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
LOL, you are a mod; congrats.
)
LOL, you are a mod; congrats. I had not noticed until you mentioned it.
RE: LOL, you are a mod;
)
Ha! :-)
It's just that since doing this moderating stuff, some posters have thought that I represent the project in some official sense, so I try not look like I do - so I don't misrepresent anything for any party! I am an absolute volunteer ( another continent ) with no special inside link/knowledge to the project....
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
So what does it take to
)
So what does it take to disperse workload across several machines while still maintaining a common and coherent database? So as to not have all the breakable eggs in one breakable basket....
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal