How heavy a load can Einstein sustain?
7 Aug 2005 0:18:46 UTC
Topic 189673
(moderation:
The Einstein BOINC project has been running pretty much rock solid for months as far as I can tell.
Based on the stats, it seems to have about 1/4 the load that SETI BOINC is struggling with.
I'm wondering how much of the SETI BOINC problems are a function of that heavier load compared to Einstein.
Said a different way, if a large number of folks migrate from either SETI BOINC or SETI Classic (once it shuts down) to Einstein, is this something Einstein can handle reliably? At the moment, it is pretty apparent that SETI BOINC can not -- so I suspect folks will 'vote with their CPU's' and at least join projects like Einstein (or Climate or others). That *might* mean a significant bump up in workload for this project.
Currently the growth here has been modest (100 to 200 users a day) while over in SETI BOINC the growth in terms of users has been a LOT higher - 10 times as much (as folks migrate from SETI classic).
So I'm wondering, if Einstein were to see say 500 to a 1000 new users a day, could it handle it?
How heavy a load can Einstein sustain?
)
Because of the shorter deadlines and longer crunch times, I suspect a much smaller percentage will migrate to Einstein than some of the other projects.
If i'm wrong I guess we will soon find out as to wether Einstein can handle it or not.
visit boinc@australia
RE: Because of the shorter
)
Could be -- folks might try Climate -- which has the much longer crunch times but much longer deadlines and offers the interim (trickle) crediting.
Berry – All-in-all, two
)
Berry – All-in-all, two different groups of crunchers, so when S@H closes down, I think there will be an influx of new members, which may cause a strain on the E@H servers, but it will perhaps be a strain that will go away rather quickly in 7 to 10 days. Hopefully, E@H will retain the graphically color blind, non-egoistical credit crunchers who either believe or disbelieve in gravitational waves.
Just an observation on my behalf, when attempting to understand why an individual would even want to join a BOINC project or similarly related project(s). I would like to feel that there exists within these project’s subscribers, a personal belief in the goal of what the project is attempting to either analyze or uncover; that is beyond just simply racking-up “ego serving figures of merit” or displaying for friends and family members a cosmetically well designed screen saver.
Furthermore, I feel, perhaps wrongly so, E@H’s data contributors have a more dedicated commitment to unlocking the mysteries of the universe via gravitational wave exploration. E@H’s drop-out rate is high, probably higher than S@H but consider some of the top performers (yourself with nearly 120,000 credits) and how much they/you have accomplished within a relatively short time span.
When viewing E@H’s “Science” board and reading some of the highly technical postings by notable contributing authors, who are also data crunching on this E@H endeavor, there’s really no comparison between the two participating memberships.
Berry, I’m sure there are many knowledgably and sincere S@H’ers who are donating their services to the advancement of science, but I would also like to feel that anyone who is enrolled in these distributed processing projects is also doing so in an e-learning frame of mind.
Theory of Gravitational Waves & LIGO
Laser Interferometer Space Antenna - LISA
JPL-Caltech
You are probably right --
)
You are probably right -- certainly with the Seti Classic folks -- as to some degree that product has elements of 'mass market' appeal to it. I think it is less the case with ANY BOINC project -- more work is involved and the 'credit boast' effort is somewhat more obscure generally.
For me, I have a large farm of workstations including some I've deployed at client sites (with the permission of the clients). I'm in my 50's and started reading science fiction back 'in the day' -- as in maybe 45 years ago -- so I've got a long history of wanting to do/know science. Heck, my brother is a physics person at NRL -- made it a life long career (PhD out of Georgetown when I was still in college).
It is science in general for me -- which is why I've added the Climate project to my BOINC group (that and the fact that both Einstein and Climate run much more reliably than SETI).