Why classic SETI@home is closing down and other facts of life.


log in

Advanced search

Message boards : Number crunching : Why classic SETI@home is closing down and other facts of life.

Previous · 1 · 2 · 3 · 4 · 5 · 6 · 7 . . . 15 · Next
Author Message
snorrre
Volunteer tester
Avatar
Send message
Joined: 7 Feb 00
Posts: 46
Credit: 16,393
RAC: 0
Norway
Message 211144 - Posted: 12 Dec 2005, 4:26:59 UTC - in response to Message 211042.


Hello
after 4 days awaiting that seti@home resolves by the appropriate way the problems related with your servers i'm displeased to announce that after a lot expending of power consumption and some money to acondicionate 7 servers to run properly the seti units processing i decide cancel my personal project due to deception that supose seti's best collaborative project i know ever, wasn't capable to preview the necessity of a High Availavility server or cluster that can't permit this bad moment and situation.

What deception? S@H has never promissed to have work available all the time. S@H has never promissed to work all of the time. The problems will be fixed eventually.

I'll have to agree on this. S@H actually tells you how smart it is to connect to multiple projects, so you can be sure your computer always is doing something good!

I hope some of my cpu-time can help some of the other BOINC-projects to find a cure or two for deadly decieses... For me, finding ET comes second. (Sorry to say that, SETI, but thats the way it is.) I'm still attached to the project, since I like to crunch for as many "good causes" as possible.
____________
Greetings from Norway.
Crunching is fun!

Matt Draper
Volunteer tester
Send message
Joined: 19 Feb 01
Posts: 69
Credit: 63,560
RAC: 74
United States
Message 211225 - Posted: 12 Dec 2005, 7:09:20 UTC

Ok. I can't take it anymore. All I hear is whine, whine, whine with little or no end in sight. So, here are facts. These are straight up, hard-core facts that apply to every facette of life and therefore are something that all can relate to. So, if you nay-sayers have any problems with these, you've got problems.

1. S@H is a BACKGROUND project. If your main goal in life is to make background objects the main objects, you're missing out on a lot of things you shouldn't. Let Seti run in the background the way it was intended. If it stops running for whatever reason, GET SOMETHING THAT DOES UNTIL SETI STARTS AGAIN!

2. Yes, there are problems. It's not BOINC. I am running 6 other projects just fine on BOINC. That's right, the program works fine. Try it, you'll see. It's not Seti. Seti is the science. It's not the people. They're working to solve the problems the uneducated are complaining about. So, here's what I propose. Let's let the un/under-paid people at Berkeley do their work with as little hassle as possible so that we can get this underfunded program as much help as it needs. When the program goes from experimental to commercial, then you can complain about things. Until then, (bluntly) SHUT UP!

3. BOINC is free. Why complain about something that's free and not even supposed to intrude on computer resources? Get a grip people! Let it run! That's what this is about. And why complain about credit? It's nice to know you contributed a lot, I'll agree there. But are you getting paid for it? Reimbursed? Rewarded with material goods? No? Didn't think so. Credits are something to let you feel good about yourself. THAT'S IT! And if you want a certificate saying how many credit hours you had or how much you contributed, open Word or Wordperfect or whatever your program of choice is and MAKE ONE! You'd feel so much better about it's design. And besides, you can complain to those technicians when your printer doesn't match the colors on the screen.

4. Minimal staff + old machinery + large base = problems eventually. I would like to see 1 person out there take 6 old machines and 1 new one, throw a program on there that gets info from tens of thousands of people (multiple times a day too) and keep everything running smoothly 100% of the time. Any computer technician that IS a technician and not a wannabe will back me on this. IT DOESN'T HAPPEN THAT WAY. Come on guys. I've seen Yahoo go down. Amazon has had troubles. And Ebay, don't get me started on those stories. And they're COMMERCIAL giants. They shouldn't have one problem right? If you think the answer's yes, turn off your computer and throw it out the window. You're not allowed technology in your life anymore.

All in all, S@H is a remarkable feat of ingenuity. It threw network computing into the spotlight. And it's still and EXPERIMENT. Let's let the experiment grow. Let's let the technicians and programmers and leaders do their jobs. Let's let the machinery do it's job. It's old. Give it a break. Your grandpa isn't competing in the Olympics and neither are these. So, run Seti. When that stops momentarily, run predictor. Run Einstein. Run Climate. Run whatever, but make it run. Don't waste precious BACKGROUND resources while your taking a shower. Let's let Seti fix. Let's let these guys bring us new stuff to play with. Let's stop the whining about Classic vs. BOINC. They both work. Some problems are just more evident than others when your nitpicky. These worthwhile efforts can use everyone they can get. But if all you're going to do is whine about a bunch of numbers that don't get you anything but personal satisfaction, or whine about a program that works perfectly, but appears to fail because you watch it like a hawk 24/7 with only one project on it, then go away. Please don't post here. We're all a little upset and wish things worked perfectly. The only difference is, we can live with it. If you want to leave, please don't announce it. Let's not get these guys down with bad. Let's help them out with valid suggestions and praise for their behind-the-scenes efforts. And most importantly, CRUNCH! It doesn't need to be specific, just crunch! Let's make this a community of crunchers, not a community of ex's.

Matt
Defender of the Little Numbers

tgm
Volunteer tester
Send message
Joined: 2 Jun 00
Posts: 16
Credit: 83,613
RAC: 0
United States
Message 211377 - Posted: 12 Dec 2005, 12:22:23 UTC - in response to Message 209002.

7. BOINC credit, while not perfect (though we're working on that), is much more fair in that it represents actual work done, and is valid between projects which do all kinds of different work. There is no way to translate Classic credit to BOINC credit, and so this will never happen. Classic credits will be noted in a separate field in a user profile (and will be eventually sync'ed up again after Classic shuts down).


The only way this is true is if the data is gone. And if this is the case, shame on somebody. If the data is there, then a very good mathematical estimation could be derived. No, it would not be perfect, but probably no less acurate than the boinc credit system. Sure, it's lots of work and calculation (another boinc project?). Now with that said, should the existing team handle it? Probably not. But I would suggest that it is a good project for some professor on the academic side to pick up and run with. It need not be limited to Berkeley either. I would hope that there is least one professor out there with the interest in the Seti project and the appropriate background in statistics who could get it rolling. Maybe a challenge needs to be staged.

WinterKnight
Volunteer tester
Send message
Joined: 18 May 99
Posts: 8219
Credit: 21,791,161
RAC: 12,914
United Kingdom
Message 211391 - Posted: 12 Dec 2005, 12:35:48 UTC - in response to Message 211381.

7. BOINC credit, while not perfect (though we're working on that), is much more fair in that it represents actual work done, and is valid between projects which do all kinds of different work. There is no way to translate Classic credit to BOINC credit, and so this will never happen. Classic credits will be noted in a separate field in a user profile (and will be eventually sync'ed up again after Classic shuts down).


The only way this is true is if the data is gone. And if this is the case, shame on somebody. If the data is there, then a very good mathematical estimation could be derived. No, it would not be perfect, but probably no less acurate than the boinc credit system. Sure, it's lots of work and calculation (another boinc project?). Now with that said, should the existing team handle it? Probably not. But I would suggest that it is a good project for some professor on the academic side to pick up and run with. It need not be limited to Berkeley either. I would hope that there is least one professor out there with the interest in the Seti project and the appropriate background in statistics who could get it rolling. Maybe a challenge needs to be staged.


I agree, it would be very easy to estimate a approximate average and say that it's fair. Like amount of WUs x 25 = amount CS from classic.

It's only you self who can see your stats for the classic on your own account page, you can't integrate it into signature pictures etc. if you don't do it your self :o/


But if it is done there cannot be a blanket 1 classic = xx BOINC credits it can only be done if the classic scientific data is still attached to the account, and can only then be done after verifying the 50+ million units, which will be when, 5 years, never, or maybe if a scientist wants all the data.
____________
Only two things are infinite: the universe and human stupidity, and I am not sure about the former. - Albert Einstein

tgm
Volunteer tester
Send message
Joined: 2 Jun 00
Posts: 16
Credit: 83,613
RAC: 0
United States
Message 211403 - Posted: 12 Dec 2005, 12:43:02 UTC - in response to Message 211225.

2. Yes, there are problems. It's not BOINC. I am running 6 other projects just fine on BOINC. That's right, the program works fine. Try it, you'll see. It's not Seti. Seti is the science. It's not the people. They're working to solve the problems the uneducated are complaining about. So, here's what I propose. Let's let the un/under-paid people at Berkeley do their work with as little hassle as possible so that we can get this underfunded program as much help as it needs.


Actually, Boinc is a part of the problem! The communication methodology that Boinc uses is definitely contributing to the bandwidth issues. Let's slow down a minute. There are some of us out here that are somewhat crusty systems people. There is obviously an architecture problem that does not seem to be addressed. If anything, the combination of Seti and Boinc is clearly showing some of Boinc's limitations. It's a volume thing that the other projects have yet to experience. My guess is that it is NFS associated, but it is impossible to say for sure without a more detailed picture. I, for one, would enjoy the opportunity to help you people out - no charge.

Astro
Volunteer tester
Avatar
Send message
Joined: 16 Apr 02
Posts: 8026
Credit: 600,015
RAC: 0
Message 211410 - Posted: 12 Dec 2005, 12:46:23 UTC
Last modified: 12 Dec 2005, 12:52:42 UTC

I have 6200 classic wus under my belt. Since, classic sent out 6-9 results per WU instead of 4 as boinc does, then 1/2 my work was wasted and I should only really get credit for useful work. cut my total in 1/2 and use 3100 wu as the basis for this conversion. Hmmm, Maybe I cheated for some of those and shouldn't get credit at all.

OK so let's assume I didn't cheat, let's work with my approximately 3100 useful work units. What value to we assign them??? As faster and faster processors are added to the project, the "granted" credit is actually dropping. So, do we use the initially higher average granted credit or the new lower average granted credit?

Now if my original classic machines still functioned they'd be claiming more credit than anything I currently own, so Maybe I should use a higher "claimed credit" that they would have gotten if they still worked?

Then many of my wus were done with the faster earlier seti classic app and shouldn't get as much credit as my slower newer classic application.

Then what about Noisey work units??? How many should I subtract from my total for those that may not be useful??

there is NO fair way to value them and they should remain as they are exactly what they are.

Yes, you can see your classic credit in signatures, see the bottom line of my signature for my up to date classic totals.
____________

Profile Mr.Pernod
Volunteer tester
Avatar
Send message
Joined: 8 Feb 04
Posts: 350
Credit: 1,015,988
RAC: 0
Netherlands
Message 211416 - Posted: 12 Dec 2005, 12:59:00 UTC - in response to Message 211381.

[off-topic]

It's only you self who can see your stats for the classic on your own account page, you can't integrate it into signature pictures etc. if you don't do it your self :o/

Crystallize, do these numbers look familiar? ;)
SETI@home classic workunits as of 15 March 2005 30,004
SETI@home classic CPU time as of 15 March 2005 130,371 hours
Everybody can see your Classic Credits on your accountpage.
It would however be nice if they would add the Classic Credits to the list under your avatar.
[/off-topic]

Profile jimmyhua
Send message
Joined: 16 Apr 05
Posts: 97
Credit: 357,522
RAC: 0
Guam
Message 211455 - Posted: 12 Dec 2005, 13:50:45 UTC
Last modified: 12 Dec 2005, 13:55:05 UTC

I am reading between the lines here.

And I think what it boils down to is the credits.

I've never used classic before. But if I was a classic user, the biggest incentive for me to move on over to BOINC is that I can be issued credit for work done on classic.

Then it would at least look like one big happy project.

There's also some bad timing for classic guys moving over to seti/Boinc right now, I've joined Seti@BOINC in March 2005. And I think we've got the longest outtage yet since that date. Problems began in Dec 6, but the front page says the problem began in Dec 8 or 9....

That's 6 days of downtime now and counting!! The longest I've seen prior to this was 3 or 4 days.


____________

WR-HW95
Volunteer tester
Send message
Joined: 12 Jan 02
Posts: 10
Credit: 13,034,091
RAC: 0
Finland
Message 211645 - Posted: 12 Dec 2005, 17:00:18 UTC - in response to Message 211416.

[off-topic]
It's only you self who can see your stats for the classic on your own account page, you can't integrate it into signature pictures etc. if you don't do it your self :o/

Crystallize, do these numbers look familiar? ;)
SETI@home classic workunits as of 15 March 2005 30,004
SETI@home classic CPU time as of 15 March 2005 130,371 hours
Everybody can see your Classic Credits on your accountpage.
It would however be nice if they would add the Classic Credits to the list under your avatar.
[/off-topic]

I have one question about this.
I see SETI@home classic workunits 5,112 on my account page,but on my CS team page it says received results 8663. Have i missed something?

Profile Mr.Pernod
Volunteer tester
Avatar
Send message
Joined: 8 Feb 04
Posts: 350
Credit: 1,015,988
RAC: 0
Netherlands
Message 211652 - Posted: 12 Dec 2005, 17:08:59 UTC - in response to Message 211645.

[off topic]


I have one question about this.
I see SETI@home classic workunits 5,112 on my account page,but on my CS team page it says received results 8663. Have i missed something?

you most likely missed the as of 15 March 2005 bit.
that's when they last updated the SETI/BOINC accounts with the Classic Credits.
a final update will follow after the Classic project has been shut down.
[/off topic]

RockyIII
Avatar
Send message
Joined: 25 May 99
Posts: 28
Credit: 11,988,185
RAC: 7,878
United States
Message 211653 - Posted: 12 Dec 2005, 17:13:15 UTC

> And I think we've got the longest outtage yet
> since that date. Problems began in Dec 6, but
> the front page says the problem began in Dec 8 or 9....

The most recent date of uploads on the setisynergy stats site is December 4.
____________

freppmaster
Send message
Joined: 16 Oct 05
Posts: 1
Credit: 20,623
RAC: 0
Sweden
Message 211682 - Posted: 12 Dec 2005, 17:37:04 UTC

Why cant the updates be released in a torrent fashion? This would save the limited connectionspeed of the project.
Does anyone know? Atleast have it as an alternative.
____________

Rick in Seattle
Send message
Joined: 4 Jun 99
Posts: 2
Credit: 854,342
RAC: 0
United States
Message 211740 - Posted: 12 Dec 2005, 18:31:16 UTC
Last modified: 12 Dec 2005, 18:33:57 UTC

I'm in the same boat. Got the notice of Classic's total shut down. So I DL'd and had to "think" my way through getting BIONIC up and running properly. It went well including getting my old Classic work units posted into my account info on the web site. Sure, the graphics aren't as nice. But, I didn't run those any way to free up CPU time for the actual work.

Got in the "groove" with BIONIC and liked the idea of multiple work units being available so the computer didn't sit idle between Internet connections. All was running smoothly for about 10 days. THEN came the crash. Sad. Yup 15 WU's waiting for upload and out of WU's to process.

Comment:
Come on guys. SETI knew full good and well what the I/O load on the system would be with the switch to BIONIC. They've had months of semi-beta testing with a partial crew running BIONIC. They knew what the increase in load would be with the Classic shut down. Never should have had this situation happen.

Yup LONG time user: 33,000 hours since 1999. SETI Classic outages have been limited and when they did occur, the web site QUICKLY posted the 'fix' info and a time frame for resumption of service. Users were 'dumb and happy' to follow the program. This? This is silly and should never have happened.

Paul Allen is pumping millions into the new radio telescope site. SURELY he can come up with a couple of 10's of thousands to get proper servers and Internet conn's going. LOL
____________

Profile Stephen R Homer
Avatar
Send message
Joined: 12 Dec 04
Posts: 32
Credit: 23,255
RAC: 0
United States
Message 211742 - Posted: 12 Dec 2005, 18:36:26 UTC - in response to Message 209002.
Last modified: 12 Dec 2005, 18:39:20 UTC

If the nay-sayers really wanted to help the would put there money where there BIG mouths are if you look at who are complaining the most it's the folks who never give a dime to this or any project its all poor me I'm not getting the proper credit if they want mine they can have them I'm here for the long haul so get on board and send cash not complaints
____________
[FONT=president]Click Please Thank You[/FONT]

Astro
Volunteer tester
Avatar
Send message
Joined: 16 Apr 02
Posts: 8026
Credit: 600,015
RAC: 0
Message 211747 - Posted: 12 Dec 2005, 18:40:32 UTC - in response to Message 211740.

Paul Allen is pumping millions into the new radio telescope site. SURELY he can come up with a couple of 10's of thousands to get proper servers and Internet conn's going. LOL

Mr. Allen and Mr. Branson are probably having too much fun taking their new plane into space to care right now. LOL

Profile Ageless
Avatar
Send message
Joined: 9 Jun 99
Posts: 12127
Credit: 2,519,625
RAC: 353
Netherlands
Message 211762 - Posted: 12 Dec 2005, 18:56:59 UTC - in response to Message 211740.

Yup LONG time user: 33,000 hours since 1999. SETI Classic outages have been limited and when they did occur, the web site QUICKLY posted the 'fix' info and a time frame for resumption of service.

Refreshing your memory on Classic in 1999. Go read some yourself.

May 22, 1999
Because of the unexpectedly large number of users, our server has become overloaded. Clients are unable to connect and get error messages. We're getting thousands of email messages every day about this.

July 1, 1999
Due to a burst steam pipe in the vicinity of the campus-wide Domain Name Server and routers, the entire Berkeley campus spent most of the day disconnected from the Internet. Sorry for any inconvenience.

August 27, 1999
A problem with our network connection to the UCB Campus has caused a high packet-loss rate, and hence a general performance problem, over the last day or two. The problem has new been fixed.

October 19, 1999
The server was down for a scheduled outage today to add memory. Unfortunately, things didn't go as planned.

October 25, 1999
Once again the server was down for a scheduled outage to add memory.

November 2, 1999
Yesterday an unnamed film crew plugged a gigawatt of lighting equipment into one of our uninterruptible power supplies. Not unexpectedly, it was interrupted, temporarily bringing down the user database machine.

November 28, 1999
There have been several periods of server connection problems this weekend for as yet unidentified reasons.

December 1, 1999
The science database server crashed at about 3 a.m. PDT.

December 11, 1999
The Data Server automatically rebooted because of SCSI errors

December 16, 1999
On Tuesday, 12/15, we came down for what we had thought would be a 3 hour outage to rebuild the science database.

December 28, 1999
Due to a screwup elsewhere on campus, there were a series of short power outages at the lab early this morning (starting around 7:00 AM PST). We waited until we were convinced that power was stable before getting all the machines back up and running (around 10:00 AM PST).


So, that was 1999 only... I didn't even add all of the outages. Go look up 2000, 2001, 2002 and 2003 for logs. Use the link I gave and change 1999 for the next year of choice.

Then compare that to the outages here (and in the old news) and pray tell... What was limited in down time?
____________
Jord

Loving awareness is free.

Profile Borgholio
Avatar
Send message
Joined: 2 Aug 99
Posts: 648
Credit: 11,495,293
RAC: 3,404
United States
Message 211773 - Posted: 12 Dec 2005, 19:06:17 UTC - in response to Message 211762.

Yup LONG time user: 33,000 hours since 1999. SETI Classic outages have been limited and when they did occur, the web site QUICKLY posted the 'fix' info and a time frame for resumption of service.

Refreshing your memory on Classic in 1999. Go read some yourself.

May 22, 1999
Because of the unexpectedly large number of users, our server has become overloaded. Clients are unable to connect and get error messages. We're getting thousands of email messages every day about this.

July 1, 1999
Due to a burst steam pipe in the vicinity of the campus-wide Domain Name Server and routers, the entire Berkeley campus spent most of the day disconnected from the Internet. Sorry for any inconvenience.

August 27, 1999
A problem with our network connection to the UCB Campus has caused a high packet-loss rate, and hence a general performance problem, over the last day or two. The problem has new been fixed.

October 19, 1999
The server was down for a scheduled outage today to add memory. Unfortunately, things didn't go as planned.

October 25, 1999
Once again the server was down for a scheduled outage to add memory.

November 2, 1999
Yesterday an unnamed film crew plugged a gigawatt of lighting equipment into one of our uninterruptible power supplies. Not unexpectedly, it was interrupted, temporarily bringing down the user database machine.

November 28, 1999
There have been several periods of server connection problems this weekend for as yet unidentified reasons.

December 1, 1999
The science database server crashed at about 3 a.m. PDT.

December 11, 1999
The Data Server automatically rebooted because of SCSI errors

December 16, 1999
On Tuesday, 12/15, we came down for what we had thought would be a 3 hour outage to rebuild the science database.

December 28, 1999
Due to a screwup elsewhere on campus, there were a series of short power outages at the lab early this morning (starting around 7:00 AM PST). We waited until we were convinced that power was stable before getting all the machines back up and running (around 10:00 AM PST).


So, that was 1999 only... I didn't even add all of the outages. Go look up 2000, 2001, 2002 and 2003 for logs. Use the link I gave and change 1999 for the next year of choice.

Then compare that to the outages here (and in the old news) and pray tell... What was limited in down time?



Don't forget back in 1998 when things got so bad, they had to hook an old 386 running Linux up as a temporary FTP server so they could squeeze out another 10 more client downloads per second.
____________


You will be assimilated...bunghole!

RockyIII
Avatar
Send message
Joined: 25 May 99
Posts: 28
Credit: 11,988,185
RAC: 7,878
United States
Message 211814 - Posted: 12 Dec 2005, 19:53:28 UTC

Perhaps the new download server will help. Here's a photo:

http://d116.com/spud/spud.jpg



____________

Profile ChiTownDale
Send message
Joined: 18 May 99
Posts: 3
Credit: 284,989
RAC: 0
United States
Message 211982 - Posted: 12 Dec 2005, 22:21:21 UTC

After running SETI for years and accumulating over 36,600 hours of computing time, I have a pretty dumb question.
If there is no bandwidth to upload and sownload actual work, then why is there plenty of bandwidth to post messages like this and make inquiries as to personal and overall SETI progress?
After all, bandwidth is bandwidth. If Boinc is bottled up, why can't SETI divert to its own servers? It is all well and good to use BOINC as its main distribution method, but having a backup seems like a wise thing to do. Just as important systems have emergency power backup for times when they may have a power failure,it would seem that SETI ought to be able to maintain an emergency alternate server path for when BINC gets swamped. After all, I am writing to a SETI server right this minute, which could easily be diverted to service uploads and downloads when this happens.
Perhaps I am looking at things in an overly simplistic fashion.
Just a thought.
____________

Profile Borgholio
Avatar
Send message
Joined: 2 Aug 99
Posts: 648
Credit: 11,495,293
RAC: 3,404
United States
Message 211994 - Posted: 12 Dec 2005, 22:31:09 UTC - in response to Message 211982.

After running SETI for years and accumulating over 36,600 hours of computing time, I have a pretty dumb question.
If there is no bandwidth to upload and sownload actual work, then why is there plenty of bandwidth to post messages like this and make inquiries as to personal and overall SETI progress?
After all, bandwidth is bandwidth. If Boinc is bottled up, why can't SETI divert to its own servers? It is all well and good to use BOINC as its main distribution method, but having a backup seems like a wise thing to do. Just as important systems have emergency power backup for times when they may have a power failure,it would seem that SETI ought to be able to maintain an emergency alternate server path for when BINC gets swamped. After all, I am writing to a SETI server right this minute, which could easily be diverted to service uploads and downloads when this happens.
Perhaps I am looking at things in an overly simplistic fashion.
Just a thought.


The website / forums take up a very small amount of bandwidth compared to the upload / download servers...too small to make a difference. Borging the webserver wouldn't be worth it. Besides, shutting down the forums would give the complainers even more to whine about later on. :-/
____________


You will be assimilated...bunghole!

Previous · 1 · 2 · 3 · 4 · 5 · 6 · 7 . . . 15 · Next

Message boards : Number crunching : Why classic SETI@home is closing down and other facts of life.

Copyright © 2014 University of California