Message boards :
Number crunching :
Is it time for Seti Enhanced?
Message board moderation
Author | Message |
---|---|
![]() ![]() Send message Joined: 31 Jul 01 Posts: 2467 Credit: 86,146,931 RAC: 0 ![]() |
Since the influx of all the Classic members, welcome by the way, the Seti system has been stressed to the breaking point. Seems to be running on the raged edge most of the time lately. Isn’t it time we bring out the enhanced Seti cruncher? I don’t think Seti has a prayer of catching up with the current system load until they bring it out. Can Matt Lebofsky, or any other developers, make any comments on this? Boinc....Boinc....Boinc....Boinc.... |
![]() ![]() Send message Joined: 8 Sep 05 Posts: 1386 Credit: 200,389 RAC: 0 ![]() |
Since the influx of all the Classic members, welcome by the way, the Seti system has been stressed to the breaking point. Seems to be running on the raged edge most of the time lately. Isn’t it time we bring out the enhanced Seti cruncher? I don’t think Seti has a prayer of catching up with the current system load until they bring it out. Read the BOINC Database Down thread. "By faith we understand that the universe was formed at God's command, so that what is seen was not made out of what was visible". Hebrews 11.3 |
![]() ![]() Send message Joined: 31 Jul 01 Posts: 2467 Credit: 86,146,931 RAC: 0 ![]() |
Jason, I read through your thread. Even searched for the word "enhanced" in it. Found nothing of any relevance to this thread. Shifting to the "enhanced" Seti cruncher would accomplish many good things. All the client computers would be slowed down because it does more detailed analysis of the work unit. It also counts the actual "flops" while crunching the work unit which would eliminate the need for the currently useless benchmark procedure. This would bring the claimed credit for different operating systems more in line with each other. Slowing down all the client computers would reduce the load on the upload and download servers, another good thing. Then the servers at Boinc/Seti could keep up with the demand and the developers could go back to developing the Boinc program rather than fixing the overload problems as they pop up. I do however acknowledge that Boinc/Seti is the stress test for the entire system due to the higher number of client computers here. Overall, I believe it is time for the enhanced cruncher. Boinc....Boinc....Boinc....Boinc.... |
![]() ![]() Send message Joined: 8 Sep 05 Posts: 1386 Credit: 200,389 RAC: 0 ![]() |
Jason, I read through your thread. Even searched for the word "enhanced" in it. Found nothing of any relevance to this thread. Well ready yes. But they cannot just say ok lets do it. They have not even finished the Merge yet. With the issues with the servers the past 24 hours and the merge on standby until they gewt back to the lab...it is definately NOT time to implement a totally new platform. If they did, the increased problems on top of what we already have, would discourage the new BOINCers from staying on the project even more than they are now. "By faith we understand that the universe was formed at God's command, so that what is seen was not made out of what was visible". Hebrews 11.3 |
![]() Send message Joined: 3 Apr 99 Posts: 1603 Credit: 2,700,523 RAC: 0 ![]() |
Since the influx of all the Classic members, welcome by the way, the Seti system has been stressed to the breaking point. Seems to be running on the raged edge most of the time lately. Isn’t it time we bring out the enhanced Seti cruncher? I don’t think Seti has a prayer of catching up with the current system load until they bring it out. Where do you see the SETI system at breaking-point? It is a little behind, but the transitioner backlog is falling. What ever issues Berkeley may have at the moment seem small compared with some over the past 6 months, that I thought were show-stoppers, but every time the guys have pulled a trick and sorted it. Right at the moment (and over the past few days) I'm downloading work, uploading work, and getting credit granted - so I'm slightly mistified about what the issue is? ![]() |
![]() ![]() Send message Joined: 15 Apr 99 Posts: 1546 Credit: 3,438,823 RAC: 0 ![]() |
I don't see any issues either to worry about. I think before switching to "enhanced" the dev. team should sort things out and start shipping the new client after all minor things have been fixed. Anyway i'm prepared with for "enhanced". Got my own clients ready to go. ![]() Join BOINC United now! |
![]() ![]() Send message Joined: 21 Oct 99 Posts: 2246 Credit: 6,136,250 RAC: 0 ![]() |
that is Crunch3r i hope the new enhanced will be start shortly i am not allowed to tell here so much, i think we have to look and see what is going.... :) Greetings from Germany NRW Ulli ![]() ![]() |
Daniel Schaalma ![]() Send message Joined: 28 May 99 Posts: 297 Credit: 16,953,703 RAC: 0 ![]() |
I don't see any issues either to worry about. Hi Crunch3r, Will you be releasing your optimized clients for Seti_Enhanced to the public when Berkeley rolls out their official enhanced client? How much time do your optimized clients save compaired to the official enhanced client? Regards, Daniel. ![]() ![]() ![]() |
![]() Send message Joined: 2 Aug 00 Posts: 1851 Credit: 5,955,047 RAC: 0 ![]() |
Maybe they oughta just go ahead and release SetiEnhanced. This would reduce a lot of message traffic and begin to shorten the database. But- I don't know anything about how the transition itself would impact the current state of all the data and servers at Berkeley. |
![]() ![]() Send message Joined: 15 Apr 99 Posts: 1546 Credit: 3,438,823 RAC: 0 ![]() |
I don't see any issues either to worry about. Hi Daniel, we are testing them at the moment. I started my first try today on my A64 3200+. so far after 2h30min it's at 30% but that's my first wu i'll have to see what happens when it's done. P.S. You got my e-mail adress :) ![]() Join BOINC United now! |
![]() Send message Joined: 21 Apr 00 Posts: 1459 Credit: 58,485 RAC: 0 ![]() |
Maybe they oughta just go ahead and release SetiEnhanced. This would reduce a lot of message traffic and begin to shorten the database. this would actually put more load on the system, due to the requirements that need to be met for a new app (lots of new entries in the database basically) so the DB would get bigger, and more stressed, and if by "message traffic" you mean on the fora, well, that's not a whole lot, considering most people don't post, i never did for classic, only started posting here after about 6 months of doing BOINC as i was interested in what was going on but lets keep the big picture in focus here, they need to get a stable version of the enhanced app in the production system, a splitter or 2 for enhanced (which will need writing, compiling and testing before production release) before we'll be able to do anything related to "enhanced" |
![]() Send message Joined: 3 Apr 99 Posts: 1603 Credit: 2,700,523 RAC: 0 ![]() |
Maybe they oughta just go ahead and release SetiEnhanced. This would reduce a lot of message traffic and begin to shorten the database. I understood that the enhanced application processes exactly the same WUs as the current app - just the enhanced application looks deeper into the data. If that is so, the no new splitters need to be set-up. Nor can I see the need for significant database load changes. The biggest single load I can see is 700,000 hosts downloading 1 or 2 meg of new application. ![]() |
Astro ![]() Send message Joined: 16 Apr 02 Posts: 8026 Credit: 600,015 RAC: 0 |
As far as I know they just have to change the 4.18 tag to the 5.XX tag. A slight change to the splitters. ![]() |
![]() ![]() Send message Joined: 15 Apr 99 Posts: 1546 Credit: 3,438,823 RAC: 0 ![]() |
but lets keep the big picture in focus here, they need to get a stable version of the enhanced app And there's a lot of work to be done to get this working properly, special case is the memory allocation malloc_a.cpp which is buggy and won't even compile using intels icc etc. But that's not of my concern. To be onest with you all i think there's lot to do. ![]() Join BOINC United now! |
![]() ![]() Send message Joined: 21 Oct 99 Posts: 2246 Credit: 6,136,250 RAC: 0 ![]() |
that are things that no one understand i tried the optimised Client on Seti enhanced with my old Server Windows 2000 Server SP4 and an XP 2.500+ Memory is going from 0 to http://home.teleos-web.de/ubrinkschmidt/screen/enserver1.jpg but after this is running fine of the usage of 64 MB... Greetings from Germany NRW Ulli ![]() ![]() |
![]() Send message Joined: 21 Apr 00 Posts: 1459 Credit: 58,485 RAC: 0 ![]() |
I understood that the enhanced application processes exactly the same WUs as the current app - just the enhanced application looks deeper into the data. If that is so, the no new splitters need to be set-up.it's the same data, but different splitters will be needed to generate the different header(s) needed to tell clients to use the "enhanced" app rather than the "standard" app, same data being processed, but new splitters will be needed (well you could change the old ones, but i doubt it, as they need 7 to keep up with the rate of work at present, and implementing new ones is easier and safer because it will affect the rest of the production system less than trying to change existing splitters, it's just good practice) Nor can I see the need for significant database load changes.well, i wasn't thinking about load in bandwidth terms, merely as DB load, which will increase at first because of all the stuff that needs to be done to add a new app to the system, and all the requests by various parts of the back-end etc. and more varied requests are (from the little i know) served slower than similar requests, especially if some form of caching is used, but my knowledge of DBs isn't vast, so any DB admin's/guru's can correct me here but yes, it will die off eventually as the WU throughput rate will drop (but it'll be helped by people not hitting the "update/retry" button every 2 seconds!) |
![]() ![]() Send message Joined: 9 Jan 00 Posts: 2562 Credit: 12,301,681 RAC: 0 ![]() |
Evening Everyone Everyone has questions and a few are over in Seti Beta trying to help insure that it is ready... I have one machine doing Seti, Einstein and Seti Enhanced... and two that are currently only doing Seti Enhanced. I was a bit of a Rabble Rouser when I did a Thread here about Seti Enhanced - Almost Perfect... The purpose was to get people here asking questions to Go LOOK! In some cases get more involved, it is Your Chance to help insure the next version has fewer problems... As there are few people really testing it makes things tougher... If You do not know if Your computer is up to Seti Enhanced Go Start Looking. Then "You" can decide without someone telling you the "Medicine Tastes Like Honey," really it does! Number Crunching Forum... Generally I have three machines Crunching. PII 400 - 719,839.67 seconds PII 400 PIII 550 - 523,069.37 seconds PIII 550 AMD64 3200 - 113,153.73 seconds AMD64 3200 If You never Go Look, then you will be surprised... No Credits for Seti Enhanced do not count here... Your Peace of Mind does R/ Al Please consider a Donation to the Seti Project. |
![]() Send message Joined: 3 Apr 99 Posts: 98 Credit: 2,667,122 RAC: 1 ![]() |
is there any free C/C++ compiler that can compile the source code? such as DJGPP,mingw,etc.. thanks!! |
![]() Send message Joined: 28 Nov 99 Posts: 402 Credit: 528,725 RAC: 0 ![]() |
is there any free C/C++ compiler that can compile the source code? I don't think setienhanced will stave off user overload too long. Computers are still getting faster and they did this a little bit before. They added more calculations. |
![]() Send message Joined: 21 Apr 00 Posts: 1459 Credit: 58,485 RAC: 0 ![]() |
I don't think setienhanced will stave off user overload too long. Computers are still getting faster and they did this a little bit before. They added more calculations. true, but enhanced takes something like 10 times as long per WU, so it'll help also computers will take quite a while to get 10 times faster, and as nice as it would be, i'm sure SETI won't get 10 times as many hosts over night ;) |
©2025 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.