no work from project AGAIN!!!


log in

Advanced search

Message boards : Number crunching : no work from project AGAIN!!!

Previous · 1 · 2 · 3 · 4 · 5 · 6 · 7 . . . 11 · Next
Author Message
ja
Send message
Joined: 20 Dec 01
Posts: 1
Credit: 201
RAC: 0
United Arab Emirates
Message 10276 - Posted: 20 Jul 2004, 11:33:04 UTC - in response to Message 10214.

> Let us see here, the beta project when computing all out was only sending out
> 20,000-30,000 results a day. Now we are sending out 600,000 results a day for
> the public project.

> In the future, when trying to compare loads, I would suggest you actually try
> comparing number of hosts instead, since it is the hosts that actually
> generate the load.
>
> Ill grant you that the bulk of the consumer space uses only one or two
> machines, but we do have a sizable part of the consumer space that has 10 or
> more machines that they are contributing to the effort. A few of them even
> have 100+ machines.
>
> ----- Rom
> BOINC Development Team, U.C. Berkeley
>
>
Agreed, but Rom, BOINC developers were warned incessantly during Alpha and Beta about the need to provide external caching facilities like SetiQueue and SetGate for S@Hv1. How can you hope to carry this load centrally. Hopefully people like Mike Ober will come to your rescue, and mine so that I can get my "farm" active again instead of just one test machine.

Heffed
Volunteer tester
Send message
Joined: 19 Mar 02
Posts: 1856
Credit: 40,736
RAC: 0
United States
Message 10279 - Posted: 20 Jul 2004, 11:35:27 UTC - in response to Message 10273.

> It doesn't matter if you WERE doing 2 million a day. When this piece of crap
> stopped, it stopped for EVERYONE which means the number yesterday is going to
> be a lot more like zero. Besides, I'm sure the classic numbers would be quite
> higher had BOINC not taken so many users and wasted their time and energy
> which COULD have been used utilizing software that ACTUALLY WORKS.

The software works. It's the servers that are having the issues...

<a> [/url]

Heffed
Volunteer tester
Send message
Joined: 19 Mar 02
Posts: 1856
Credit: 40,736
RAC: 0
United States
Message 10281 - Posted: 20 Jul 2004, 11:42:43 UTC - in response to Message 10276.

> Agreed, but Rom, BOINC developers were warned incessantly during Alpha and
> Beta about the need to provide external caching facilities like SetiQueue and
> SetGate for S@Hv1. How can you hope to carry this load centrally. Hopefully
> people like Mike Ober will come to your rescue, and mine so that I can get my
> "farm" active again instead of just one test machine.

Errrmmm.... So, where do these external caching facilities get their WUs? ;-)

The whole point of such a thing is more or less moot with BOINC. Since the average deadline is two weeks, a cache larger than two weeks is simply stupid. I fail to see the relevance of whether BOINC is doing the caching, or an application like SetiQueue. A cache is a cache, and the WUs must come from Berkeley.

Profile slavko.sk
Avatar
Send message
Joined: 27 Jun 00
Posts: 346
Credit: 408,100
RAC: 0
Slovakia
Message 10286 - Posted: 20 Jul 2004, 12:08:30 UTC - in response to Message 10281.

I think that it is so clear that I don't understand complaining about it.
Lot of people are swithing to BOINC version and every one is caching WU for week or more in advance. So it produce heavy load to Berkeley's servers.
On other side, SETI@BOINC is STILL in beta test phase, so we *HAVE TO* count with some outages. Like today, my PC was without worj for 30 minutes, not so bad.
But I *HAVE TO* say that last 2 week it works without any my touch, I was on vacation 2000km from here, I set cache to 3-6 days, I did not cache for 2 weeks or so. And take a look on PC use: 99.9% which means that is was working nearly all the time. And PC was 2 weeks closed in my office alone, without ANY interaction from anybody. THAT's NICE.
Isn't?

> > Agreed, but Rom, BOINC developers were warned incessantly during Alpha
> and
> > Beta about the need to provide external caching facilities like SetiQueue
> and
> > SetGate for S@Hv1. How can you hope to carry this load centrally.
> Hopefully
> > people like Mike Ober will come to your rescue, and mine so that I can
> get my
> > "farm" active again instead of just one test machine.
>
> Errrmmm.... So, where do these external caching facilities get their WUs? ;-)
>
> The whole point of such a thing is more or less moot with BOINC. Since the
> average deadline is two weeks, a cache larger than two weeks is simply stupid.
> I fail to see the relevance of whether BOINC is doing the caching, or an
> application like SetiQueue. A cache is a cache, and the WUs must come from
> Berkeley.
>
>



Try WAP user info http://slavko.sk/wap/seti

Daniel Schaalma
Volunteer tester
Avatar
Send message
Joined: 28 May 99
Posts: 297
Credit: 16,953,703
RAC: 0
United States
Message 10292 - Posted: 20 Jul 2004, 12:32:04 UTC
Last modified: 20 Jul 2004, 12:33:50 UTC

> Agreed, but Rom, BOINC developers were warned incessantly during Alpha and
> Beta about the need to provide external caching facilities like SetiQueue and
> SetGate for S@Hv1. How can you hope to carry this load centrally. Hopefully
> people like Mike Ober will come to your rescue, and mine so that I can get my
> "farm" active again instead of just one test machine.

The only thing that SetiQueue and SetiGate do that BOINC does not do, is allow you to cache a huge amount of WU's on ONE of YOUR machines, and then distribute them to other machines of YOURS that may not have internet access. Those SITES do not cache the WU's. All WU's come from Berkeley.

Also, I am getting really sick of people in these forums badmouthing the hardworking developers at Berkeley. They are doing everything they can to get BOINC running smoothly. They are a reletively small team with a very limited budget. I know that it is frustrating to have computers with no WU's to crunch. I am a hard core cruncher myself, and I have built 16 machines now that crunch Seti segments. But seriously, people... Step back, take a deep breath, and remember that this is first and foremost a scientific project that you have VOLUNTEERED to assist. No one is FORCING you to do this. Things will get settled out in time. Please be patient and let the developers make it work. If you have a complaint, please post it in a civilized mannor. REMEMBER, there are school children doing this at home and at school for science projects, and some of you are not setting a very good example for those youngsters.

Regards, Daniel.

Profile Christopher Hauber
Avatar
Send message
Joined: 10 Feb 01
Posts: 196
Credit: 71,611
RAC: 0
United States
Message 10304 - Posted: 20 Jul 2004, 13:14:23 UTC - in response to Message 10273.
Last modified: 20 Jul 2004, 13:21:30 UTC

Ok Brainsmasher, since it seems someone has smashed your brain, I'm going to make this simple for you. Your post has no merit. If you think you know better, then set up the project yourself. Otherwise, stop your bickering and grow up and get some realistic expectations. Rom told us what is going on and that is the end of it. If you don't like it, tough.

Chris

EDIT -- I missed one part of your posts about SETI Classic numbers the first time around. Making SETI classic numbers higher is irrelavent considering that BOINC is going to be able to do more calculations than SETI classic (and may already have some of them implemented) and considering the same workunits were being crunched over and over and over and over an.. well I hope you get the idea. There just wasn't really much new data being processed.


> "Now for some perspective here, since we are being compared to SAH v1's
> 137,000 WU's processed a day, now lets compare that to this projects 200,000
> WU's processed a day before things became unbalanced. "
>
> It doesn't matter if you WERE doing 2 million a day. When this piece of crap
> stopped, it stopped for EVERYONE which means the number yesterday is going to
> be a lot more like zero. Besides, I'm sure the classic numbers would be quite
> higher had BOINC not taken so many users and wasted their time and energy
> which COULD have been used utilizing software that ACTUALLY WORKS.
>
> If you want to make a comparison, you compare two like objects. Like
> the returns in a 24 hour period. Not the classic returns of one day compared
> to the returns one day LAST WEEK for BOINC.
>
> Get your s**t up to speed and I not only will listen, I'll switch back to
> BOINC (notice I am still here after uninstalling that piece of
> crap)......but as long as it a NON WORKING piece of garbage....toot your horn
> to the folks more worried about their screensaver than returning WU's.
>

Profile Christopher Hauber
Avatar
Send message
Joined: 10 Feb 01
Posts: 196
Credit: 71,611
RAC: 0
United States
Message 10317 - Posted: 20 Jul 2004, 13:52:44 UTC - in response to Message 10292.

More voices of reason! I love it :)

Using third party caching is a pretty moot point with BOINC. Two weeks worth of cache is a pretty substantial amount. Even a week is enough to get through just about any shortage, at least until the project is no longer able to keep up with the processing on a permanent basis. That WILL happen too, especially with the increasing speeds of computers.

I'm not as familiar with other operating systems as I am with the Windows ones, but I would expect there to be some way to provide internet access through the "one computer with internet access" with other OSs besides Windows, especially if they are networked which I would have to assume if the queuing program is able to distribute workunits to them.

If I'm wrong then I can understand it to a point, but with deadlines I see that as a more hazardous endeavor. Otherwise, I don't see any problem with BOINC handling all of the caching.

Chris

> The only thing that SetiQueue and SetiGate do that BOINC does not do, is allow
> you to cache a huge amount of WU's on ONE of YOUR machines, and then
> distribute them to other machines of YOURS that may not have internet access.
> Those SITES do not cache the WU's. All WU's come from Berkeley.
>
> Also, I am getting really sick of people in these forums badmouthing the
> hardworking developers at Berkeley. They are doing everything they can to get
> BOINC running smoothly. They are a reletively small team with a very limited
> budget. I know that it is frustrating to have computers with no WU's to
> crunch. I am a hard core cruncher myself, and I have built 16 machines now
> that crunch Seti segments. But seriously, people... Step back, take a deep
> breath, and remember that this is first and foremost a scientific project that
> you have VOLUNTEERED to assist. No one is FORCING you to do this. Things will
> get settled out in time. Please be patient and let the developers make it
> work. If you have a complaint, please post it in a civilized mannor. REMEMBER,
> there are school children doing this at home and at school for science
> projects, and some of you are not setting a very good example for those
> youngsters.
>
> Regards, Daniel.
>

STE\/E [BlackOps]
Volunteer tester
Send message
Joined: 29 Mar 03
Posts: 1137
Credit: 3,041,146
RAC: 0
United States
Message 10324 - Posted: 20 Jul 2004, 14:20:02 UTC

How much are the BOINC Dev's paying you Chris, it's either that or I have to get whatever you are smoking. I've never seen anybody so enthusiastic about something thats a piece of crap 95% of the time...

Profile Liberto [Valencia]
Avatar
Send message
Joined: 24 Jul 01
Posts: 131
Credit: 29,008
RAC: 0
Spain
Message 10331 - Posted: 20 Jul 2004, 14:32:19 UTC

I do agree with some of the people whom are commenting here that we have to be patient and let the guys in Berkeley work their way out of the problems.

Patience is a virtue and not many people are showing it. If you would not consider this as an Olympic competition and instead of putting up so many CPU's for a single user, everybody stay with one and then you would receive enough units to analize.

As some say, nobody forces us to stay here. Stop your CPU's, put nice music and rest. That will probably calm your nerves down.

Cheers to the guys in Berkeley for their best efforts to keep this working.

Thanks Ron Walton and the rest of the team!!!

Profile Christopher Hauber
Avatar
Send message
Joined: 10 Feb 01
Posts: 196
Credit: 71,611
RAC: 0
United States
Message 10344 - Posted: 20 Jul 2004, 15:11:37 UTC - in response to Message 10324.
Last modified: 20 Jul 2004, 15:20:19 UTC

I wish I was paid. The fact of the matter though is I have a little thing called patience. That and I try to think of things from the point of view of the developers rather than just an end user. I didn't pay for the software. So there really isn't a monetary investment. Ok, electricity. But if I'm out of workunits, no more electricity is used than would be used otherwise. So the only cost is incurred when it is actually doing its job. It doesn't really take any time. The computer handles all of the processing. That's not my time. So all that comes out of my time is a few minutes to download and set up if that. Who cares? I spend more time brushing my teeth in the morning. The message boards take a fair amount of time, but that's just something I choose to do to find out more about how things work and help other people. Ok, so no time invested. No money invested. What great loss do I have to complain about? It's hardly an interactive program so it certainly can't be that it's hard to use (unless you have an ISA proxy and then there is a bit more challenge to setting it up, but there are decent instructions on the boards for that). And for some reason, I just can't say it's crap 95% of the time. I don't know why that is.. Oh! Maybe that's because it downloads workunits when they are available (retries periodically if they aren't), processes them, returns them (when it is able), and then compares my results with others to ensure accuracy and grants me credit for crunching them, just like it is supposed to. Things don't always happen efficiently because this component flakes out now, and another one then, but it happens. And the outages thusfar seem to have only caused delays and not losses, so even better. Things have been running fairly smoothly for a while with the exception of yesterday. And units are still a bit hard to come by, but I get one or two every once in a while. From the looks of things though the system is just playing a little bit of catchup while people fill their cache again. Once the system is caught up and people's computers filled up, things will be basically normal again.

The problem is, you just don't look at all of the aspects that influence how the system works. The caching system makes it harder to initially make workunits available (or after a delay in producing them) because instead of just a few people stockpiling and having to refill their supplies, everybody's computer is doing that.

Chris

EDIT -- One last thing. One reason for being so "enthusiastic" as you put it, is that almost every complaint about how "bad" everything is has been in ignorance, some far worse than others, and I can't stand how people can sit there and complain so poorly and immaturely about when they have no stake in the project to complain about and when they either no nothing about what they are talking about or just haven't thought things through.

> How much are the BOINC Dev's paying you Chris, it's either that or I have to
> get whatever you are smoking. I've never seen anybody so enthusiastic about
> something thats a piece of crap 95% of the time...
>

Profile Marcin Radzikowski
Send message
Joined: 15 May 99
Posts: 17
Credit: 5,279,985
RAC: 3,671
Canada
Message 10348 - Posted: 20 Jul 2004, 15:22:22 UTC

If you really want your machine to "crunch data" why not sign up for other BOINC projects? There is Predictor@home and soon Einstein@home.

Profile Marcin Radzikowski
Send message
Joined: 15 May 99
Posts: 17
Credit: 5,279,985
RAC: 3,671
Canada
Message 10349 - Posted: 20 Jul 2004, 15:23:04 UTC

Oh yeah...Predictor@home has something like 6000 users...they need some help!

Wolfmarkt
Send message
Joined: 16 May 99
Posts: 5
Credit: 584,461
RAC: 701
United States
Message 10355 - Posted: 20 Jul 2004, 15:32:12 UTC
Last modified: 20 Jul 2004, 16:59:39 UTC

For something that is supposed to quietly run in the background, unobtrusively using idle time that would otherwise go wasted and doing a good job of it, it amazes me that there is so much static regarding something as simple as having or not having work units. There is not one single thing wrong with not having work units to process, BOINC quite elegantly sits and waits for awhile before again requesting more work. I'd bet that were the tray icon not equipped to notify when there are messages to be read many people wouldn't even know it was idle.

I think we "volunteers" should relax and let the authors of this "free" "beta" "scientific" software work through the problems and stop worrying so much about all the apparently precious cycles BOINC is not using when it has nothing to do and instead, use our own truly precious, idle brain's cycles pursuing greater things.

Anonymous
Send message
Joined: 15 Jan 02
Posts: 307
Credit: 24,137
RAC: 0
Afghanistan
Message 10377 - Posted: 20 Jul 2004, 16:48:12 UTC
Last modified: 20 Jul 2004, 16:56:56 UTC

"to keep your comps crunching you're are going to have to develop a MULTI Project attitude."

If you were using SETI classic INSTEAD of BOINC you'd KNOW that is a false statement.

"If you really want your machine to "crunch data" why not sign up for other BOINC projects? There is Predictor@home and soon Einstein@home."

I don't want to "crunch data".....I want to find aliens.....BIG difference.

"The software works. It's the servers that are having the issues..."

SETI classic works just fine.....so in essence IT IS BOINC that was causing me problems.

"If you think you know better, then set up the project yourself. Otherwise, stop your bickering and grow up and get some realistic expectations."

I don't have to....it's what SETI "classic " is for......and in America I have every right to express my opinion about crappy software. As for my expectations....I don't think software that WORKS is an unrealistic expectation.....especially since SETI classic does just that...IT WORKS.

"Making SETI classic numbers higher is irrelavent considering that BOINC is going to be able to do more calculations than SETI classic (and may already have some of them implemented)"

PURE SPECULATION. My expirence with BOINC shows that it can't even connect to get WU's. Therefore I certainly don't see how you "think" it's going to DO more calculations.....but you're the "genius" here, right?

</img>

Profile Marcin Radzikowski
Send message
Joined: 15 May 99
Posts: 17
Credit: 5,279,985
RAC: 3,671
Canada
Message 10379 - Posted: 20 Jul 2004, 17:01:43 UTC

BrainSmashR:

So, how much have you contributed to the SETI project? And, I don't mean in the number of SETI@home work units processed...I mean in $$$. The fact of the matter is that SETI@home is a scientific project of immense magnitude being developed and managed by a handful of people. So, if you EXPECT software that works you'd better pay for it, otherwise let the SETI people do their work to get the system running. BOINC is completely different from SETI classic and the analysis software is not crap and it works. There are teething problems with the WU distribution. If you'd been part of SETI@home since the beginning you'd know that there were some major outages at the start and even well into the project.

Wolfmarkt
Send message
Joined: 16 May 99
Posts: 5
Credit: 584,461
RAC: 701
United States
Message 10380 - Posted: 20 Jul 2004, 17:07:24 UTC

I'm beginning to think that this quote is quite appropriate for this thread:

"Sometimes I think scientists should quit searching for intelligent life in outer space and see if they can find any on this planet." -- Charley Reese, journalist, newspaper columnist

Profile Christopher Hauber
Avatar
Send message
Joined: 10 Feb 01
Posts: 196
Credit: 71,611
RAC: 0
United States
Message 10384 - Posted: 20 Jul 2004, 17:31:38 UTC - in response to Message 10377.

That's not a false statement. It is precisely true. SETI classic is going to be turned off in the near future. At that point, the verification process of BOINC will eliminate a lot of the overprocessing currently done by classic, and cause the work supply to be exhausted much faster. And with increasingly fast computers, you can count on the fact that there will not be enough data to provide a continuous supply of work to everyone who wants to process data. BOINC is a good answer to that, by providing other uses for your computer with the same program when data is unavailable. If you don't want to process anything else, fine. You don't have to. But don't whine about it. SETI@Home is not really looking for intelligent life on other planets so much as it is looking for SIGNS (why do you think they call signals that are found candidate signals?) of life on other planets. The project may not turn up any results, the results may not be in your lifetime or mine. Even in the event a good candidate is found and intelligent life is the only concievable source, it is extremely unlikely that anything can be done in our lifetimes about it. SETI isn't transmitting, it's listening. And any transmissions made take years to even get to the closest star. And if your theory about aliens coming here and basically taking over is true, what are you going to do about it?

You are worrying about things that you have absolutely no control over whatsoever and never will. So relax and take things in stride. It is all you can do.

SETI Classic doesn't give you realistic expectations. If it did, you would have a lot more sense in your head and wouldn't be griping about the normal course of things. I wasn't here when SETI Classic was released, but from what I have read, it went through months of problems when it was released as well. Yet you put all your stock in it anyway. And do NOT give me that load of bull about being in America and having the right to express your opinion. That's not what this is about. This is about you making mostly unfounded accusations based on a lack of knowledge, experience, patience, and understanding. Sure, you have the "right" to complain, but you're only making yourself look unintelligent and narrow minded. If you want to use that line about being free to do whatever because you are an American, try being a little more open minded first so you don't sound so arogant and defiant. Your nationality also has very little relavence here on the page of an international internet scientific research project. None of that implies that your country has any special rights over anyone else. Here everyone has just as much right as anyone else. I don't care where you are from, I just expect people here to try and use their brains. That's what you do in science.

And for the record, BOINC does work. It works for me. It works for most people around here. The servers sometimes have temporary outages and glitches, but it all works. Don't tell me it doesn't work. I don't care much about credit, but I have about 7000 granted credits that say you are flat out wrong to say it doesn't work. Over a thousand other people have more than that.

Again, I have over 7000 granted credits. Others have more. We are getting workunits or we wouldn't be getting credits. Workunits are a bit hard to come by at the moment, but they are there. And what makes me think it will do more calculations? Because I can read. BOINC is not SETI@Home. SETI@Home is essentially a core program that processes workunits, and BOINC controls that program. During the life of SETI Classic, the computational abilities were changed and updated to look for different types of signals and patterns. It is entirely possible that more modifications have been made for use with BOINC. And I doubt it has been done yet, but one reason for going to BOINC is to be able to seamlessly add the ability process data recorded at other locations as they become available and to process the full frequency range currently being recorded instead of just a narrow band. Read. You will find lots of helpful information if you just read.

Chris

> "to keep your comps crunching you're are going to have to develop a MULTI
> Project attitude."
>
> If you were using SETI classic INSTEAD of BOINC you'd KNOW that is a false
> statement.
>
> "If you really want your machine to "crunch data" why not sign up for other
> BOINC projects? There is Predictor@home and soon Einstein@home."
>
> I don't want to "crunch data".....I want to find aliens.....BIG difference.
>
> "The software works. It's the servers that are having the issues..."
>
> SETI classic works just fine.....so in essence IT IS BOINC that was causing me
> problems.
>
> "If you think you know better, then set up the project yourself. Otherwise,
> stop your bickering and grow up and get some realistic expectations."
>
> I don't have to....it's what SETI "classic " is for......and in America I have
> every right to express my opinion about crappy software. As for my
> expectations....I don't think software that WORKS is an unrealistic
> expectation.....especially since SETI classic does just that...IT WORKS.
>
> "Making SETI classic numbers higher is irrelavent considering that BOINC is
> going to be able to do more calculations than SETI classic (and may already
> have some of them implemented)"
>
> PURE SPECULATION. My expirence with BOINC shows that it can't even connect to
> get WU's. Therefore I certainly don't see how you "think" it's going to DO
> more calculations.....but you're the "genius" here, right?
>
>

Profile Christopher Hauber
Avatar
Send message
Joined: 10 Feb 01
Posts: 196
Credit: 71,611
RAC: 0
United States
Message 10385 - Posted: 20 Jul 2004, 17:35:27 UTC - in response to Message 10380.

Agreed. I didn't know the quote itself, but I've had that thought in my mind for a while now.

> I'm beginning to think that this quote is quite appropriate for this thread:
>
> "Sometimes I think scientists should quit searching for intelligent life in
> outer space and see if they can find any on this planet." -- Charley Reese,
> journalist, newspaper columnist
>
>

CyberGoyle
Avatar
Send message
Joined: 2 Jun 99
Posts: 158
Credit: 1,710,376
RAC: 596
United States
Message 10388 - Posted: 20 Jul 2004, 17:45:39 UTC

First off, thanks Christopher for saying the things I want to say but hadn't found the right words to type. Second, lets all give up on this thread - its a waste of time. If Berkeley gets to keep all the '2 cents' people throw in, they will be able to fund the Seti project until about 99 years after the sun burns out.


<a>

Profile Darth Dogbytes™
Volunteer tester
Send message
Joined: 30 Jul 03
Posts: 7512
Credit: 2,021,148
RAC: 0
United States
Message 10389 - Posted: 20 Jul 2004, 17:46:20 UTC

Well stated Chris; to the chronic complainers-
____________
Account frozen...

Previous · 1 · 2 · 3 · 4 · 5 · 6 · 7 . . . 11 · Next

Message boards : Number crunching : no work from project AGAIN!!!

Copyright © 2014 University of California