Keep enough work on disk to last between???

Message boards : Number crunching : Keep enough work on disk to last between???
Message board moderation

To post messages, you must log in.

1 · 2 · Next

AuthorMessage
Profile Ryvr
Volunteer tester

Send message
Joined: 27 Oct 03
Posts: 5
Credit: 578,694
RAC: 0
United States
Message 5982 - Posted: 10 Jul 2004, 7:06:28 UTC

I am not getting any S@H WU's. Wondering if I need certain magic numbers in my "Keep enough work on disk to last between"... I had .1-3, .1-5, 2.9-5, and now 4.5-5. I am always connected so that's not an issue. Can anyone share their #'s if they are working?
Peace,
Sruthan
ID: 5982 · Report as offensive
Darren
Volunteer tester
Avatar

Send message
Joined: 2 Jul 99
Posts: 259
Credit: 280,503
RAC: 0
United States
Message 5985 - Posted: 10 Jul 2004, 7:19:25 UTC - in response to Message 5982.  
Last modified: 11 Jul 2004, 6:24:19 UTC

> I am not getting any S@H WU's. Wondering if I need certain magic numbers in my
> "Keep enough work on disk to last between"... I had .1-3, .1-5, 2.9-5, and now
> 4.5-5. I am always connected so that's not an issue. Can anyone share their
> #'s if they are working?

Any of those numbers should allow you to get something. If it's not even giving you one work unit, it something other than those settings.

What message does it give you when it connects and asks for work?


Darren


ID: 5985 · Report as offensive
Profile Ryvr
Volunteer tester

Send message
Joined: 27 Oct 03
Posts: 5
Credit: 578,694
RAC: 0
United States
Message 5986 - Posted: 10 Jul 2004, 7:23:37 UTC

SETI@home - 2004-07-10 03:02:48 - Sending request to scheduler: http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi

SETI@home - 2004-07-10 03:02:51 - Scheduler RPC to http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi succeeded

That's it. There doesn't really seem to be an error... but no WU's. I have been getting plenty from Predictor. I thought the S@H servers were down, but I see on the boards lots of other people talking about getting WU's just now, so I'm confused.
Hmmm.
Peace,
Sruthan
ID: 5986 · Report as offensive
Darren
Volunteer tester
Avatar

Send message
Joined: 2 Jul 99
Posts: 259
Credit: 280,503
RAC: 0
United States
Message 5989 - Posted: 10 Jul 2004, 7:33:03 UTC - in response to Message 5986.  
Last modified: 11 Jul 2004, 6:23:59 UTC

> SETI@home - 2004-07-10 03:02:48 - Sending request to scheduler:
> http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi
>
> SETI@home - 2004-07-10 03:02:51 - Scheduler RPC to
> http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi succeeded
>
> That's it. There doesn't really seem to be an error... but no WU's. I have
> been getting plenty from Predictor. I thought the S@H servers were down, but I
> see on the boards lots of other people talking about getting WU's just now, so
> I'm confused.

Ahh, do you have predictor units running now? If so, it won't try to get any work from anywhere until it hits the low-water mark.

It's connecting successfully, it's just not requesting anything, so likely it hasn't hit that mark yet. And if you've only got a few predictor units but just recently upped your setting to 4.5 minimum, it likely just hasn't had time to propagate throughout yet. Those settings will eventually take hold and work their way from boinc to predictor and it will start using the current settings.

But again, first thing it's got to do is hit the low-water mark. Then it should ask for work from somewhere. Depending upon your resource allocations, though, it may ask for work from predictor again.


Darren


ID: 5989 · Report as offensive
Profile Thierry Van Driessche
Volunteer tester
Avatar

Send message
Joined: 20 Aug 02
Posts: 3083
Credit: 150,096
RAC: 0
Belgium
Message 6011 - Posted: 10 Jul 2004, 9:02:44 UTC - in response to Message 5986.  
Last modified: 20 Jul 2004, 12:10:04 UTC

> SETI@home - 2004-07-10 03:02:48 - Sending request to scheduler:
> http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi
>
> SETI@home - 2004-07-10 03:02:51 - Scheduler RPC to
> http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi succeeded
>
> That's it. There doesn't really seem to be an error... but no WU's. I have
> been getting plenty from Predictor. I thought the S@H servers were down, but I
> see on the boards lots of other people talking about getting WU's just now, so
> I'm confused.

When you change your preferences, if you do an "Update" in the tab "Projects", you should have in the "Messages" tab a sentence showing the date and time the last preferences have been modified.

Greetings from Belgium.
ID: 6011 · Report as offensive
Profile Mike Special Project $75 donor
Volunteer tester
Avatar

Send message
Joined: 17 Feb 01
Posts: 34253
Credit: 79,922,639
RAC: 80
Germany
Message 6041 - Posted: 10 Jul 2004, 11:02:40 UTC

Hi

You have to finnish the work from predictor first.
You will get the message cache low watermark hit - requesting xxx seconds of work.

Then you will get it when work is available.

greetz Mike

ID: 6041 · Report as offensive
Rob

Send message
Joined: 26 Nov 99
Posts: 3
Credit: 102,856
RAC: 0
Australia
Message 6044 - Posted: 10 Jul 2004, 11:12:26 UTC

I,m having a similar problem to Sruthan. I've been running Boinc for about a week & have my resource share set at 100000 for seti@home and 0.000001 for predictor. So far I have received a grand total of either 2 or 4(cant remember exactly) seti work units at either 2 or 3 at a time. Yet I receive up to 20 predictor work units at a time. Plus I still havn't received any seti credits, while predictor credits are added within a day or two.

Rob.
ID: 6044 · Report as offensive
Darren
Volunteer tester
Avatar

Send message
Joined: 2 Jul 99
Posts: 259
Credit: 280,503
RAC: 0
United States
Message 6132 - Posted: 10 Jul 2004, 14:57:08 UTC - in response to Message 6044.  
Last modified: 11 Jul 2004, 6:23:22 UTC

> I,m having a similar problem to Sruthan. I've been running Boinc for about a
> week & have my resource share set at 100000 for seti@home and 0.000001 for
> predictor. So far I have received a grand total of either 2 or 4(cant remember
> exactly) seti work units at either 2 or 3 at a time. Yet I receive up to 20
> predictor work units at a time. Plus I still havn't received any seti credits,
> while predictor credits are added within a day or two.

If you reduce your cache size, you will be requesting work more frequently (as you'll be getting less work each time you get some). This will increase your likelihood of making a request when seti actually happens to be up.

You can either set your cache low altogether (something like 0.3 minimum and 0.5 maximum) or you can just bring the 2 number closer together if the max is higher (something like 2.8 minimum and 3 maximum).


Darren


ID: 6132 · Report as offensive
Rob

Send message
Joined: 26 Nov 99
Posts: 3
Credit: 102,856
RAC: 0
Australia
Message 6378 - Posted: 11 Jul 2004, 5:21:36 UTC - in response to Message 6132.  
Last modified: 11 Jul 2004, 5:35:36 UTC

> If you reduce your cache size, you will be requesting work more frequently (as
> you'll be getting less work each time you get some). This will increase your
> likelihood of making a request when seti actually happens to be up.
>
> You can either set your cache low altogether (something like 0.3 minimum and
> 0.5 maximum) or you can just bring the 2 number closer together if the max is
> higher (something like 2.8 minimum and 3 maximum).
>
>
> Darren
>
I was just thinking this morning that the problem might be the opposite. I have the cache set at 0.1 Min and 1 Max, and since Seti seems to be handing out work units with a high completion time (20 to 30 hours I believe) they won't fit in the cache. So I've tried raising the Max to 3 days. I'll see how it goes over the next few days, and if it doesn't work I'll try your suggestion. Thanks anyway. BTW I should add that I havn,t received any Seti work units for several days, which I think coincides with the start of the large work units.

Rob
ID: 6378 · Report as offensive
Lionel

Send message
Joined: 25 Mar 00
Posts: 680
Credit: 563,640,304
RAC: 597
Australia
Message 6381 - Posted: 11 Jul 2004, 5:48:33 UTC - in response to Message 5982.  

> I am not getting any S@H WU's. Wondering if I need certain magic numbers in my
> "Keep enough work on disk to last between"... I had .1-3, .1-5, 2.9-5, and now
> 4.5-5. I am always connected so that's not an issue. Can anyone share their
> #'s if they are working?
> Peace,
> Sruthan
>
>
ID: 6381 · Report as offensive
Lionel

Send message
Joined: 25 Mar 00
Posts: 680
Credit: 563,640,304
RAC: 597
Australia
Message 6384 - Posted: 11 Jul 2004, 5:54:07 UTC - in response to Message 5982.  

> I am not getting any S@H WU's. Wondering if I need certain magic numbers in my
> "Keep enough work on disk to last between"... I had .1-3, .1-5, 2.9-5, and now
> 4.5-5. I am always connected so that's not an issue. Can anyone share their
> #'s if they are working?
> Peace,
> Sruthan
>
> Whilst it might appear as an exageration, from the posts it appears as though everyone is having problems with BOINC. I have mine set to 4.5 and 5.5 days. The report deadline seems to be a fortnight out. My suggestion is get enough work units to last a little over a week.
ID: 6384 · Report as offensive
Darren
Volunteer tester
Avatar

Send message
Joined: 2 Jul 99
Posts: 259
Credit: 280,503
RAC: 0
United States
Message 6387 - Posted: 11 Jul 2004, 6:15:31 UTC - in response to Message 6384.  


> > Whilst it might appear as an exageration, from the posts it appears as
> though everyone is having problems with BOINC. I have mine set to 4.5 and 5.5
> days. The report deadline seems to be a fortnight out.

That is correct. All the current work units have a deadline 14 days from their issue date.

> My suggestion is get
> enough work units to last a little over a week.

That's a good setting if you are only attached to seti. If you have a "backup" in the mix (such as predictor) and you want to make sure you don't get a large number of predictor units when seti happens to be down, then a large cache can be bad. If your cache has a variation from low to high of 1 day, when seti is down and you default to the backup you will get a full day's worth of work from that site. An all around low cache will have you getting and giving 1 or 2 at a time. A large cache with a narrow range from max to min will have you keeping a lot in the cache, but still trading them out in small numbers.



ID: 6387 · Report as offensive
Darren
Volunteer tester
Avatar

Send message
Joined: 2 Jul 99
Posts: 259
Credit: 280,503
RAC: 0
United States
Message 6389 - Posted: 11 Jul 2004, 6:22:44 UTC - in response to Message 6378.  


> I was just thinking this morning that the problem might be the opposite. I
> have the cache set at 0.1 Min and 1 Max, and since Seti seems to be handing
> out work units with a high completion time (20 to 30 hours I believe) they
> won't fit in the cache.

Rob,

The cache should exceed without a problem by whatever is necessary to get a full work unit. It should get how ever many work units it needs to go just over the max when it refills. With a max of 1 day and the combination of the large times reported, it would only get 1 wu then it would think it had gone over the max and stop there.

Everything indicates that the work unit times should be coming back to correct soon.

Do you have any predictor work units in your cache now? If so, remember that they count towards your times, so you may have to wait until some of them are done. What kind of message appears in your log after boinc contacts the seti scheduler. If it only reports "succeeded" after an attempt, then it simply hasn't hit the low-water mark yet and is not asking for any work. If it reports something else, post it here and we can help you clear it up.



ID: 6389 · Report as offensive
Belial

Send message
Joined: 22 Jan 02
Posts: 47
Credit: 63,100
RAC: 0
United States
Message 6406 - Posted: 11 Jul 2004, 7:28:33 UTC
Last modified: 11 Jul 2004, 7:33:21 UTC

Big caches aren't very useful for some projects.


Preditor seems to run a pretty tight schedule and needs the results fairly fast.


set your cache for 2 weeks worth of seti if you want, but your still only going to get as many as seti has or wants you to have at once time.


It's fairly obvious that with a shortage of wu's people should go ahead and join another projects, even if they don't really want to.


Expect wu shortages to be the norm rather then the exception, especially when seti@home classic shuts down. Unless they bring the other radio telescopes on line or do more extensive processing there is just too much cpu power for the work required.


It's obvious that astro-pulse needs to start up ASAP. That will give a secondary project that many seti fans would find comfortable. And since it uses the same wu's from seti to reprocess including those that seti classic has already put into the finished and over with tape.
ID: 6406 · Report as offensive
Profile Timi
Volunteer tester

Send message
Joined: 7 Oct 99
Posts: 25
Credit: 6,533,108
RAC: 0
Greece
Message 6419 - Posted: 11 Jul 2004, 8:14:58 UTC

Well, the same here.

With classic-seti-client I was returning 8-9 Dus every day with a cache of 20 DUs.
These days with the new boinc-client, I return 6-7, but with a cache of 2 DUs only!!
I have to connect through dialup 56K, every 5-6 hours to get new dus (and during the night I have to get back to old-seti) 8-(
I have select a period of 2 to 3 days to keep, through preferences, but the dus are not even 5-6, just to get over a day.
Is this has to do with bad cpu benchmarks through the boinc-client or the project is not giving more than a certain number of dus per connection/host id?

Now, the boinc-client asks for 20576 seconds of work, and it is downloading only 1 du!

Any help?
ID: 6419 · Report as offensive
Petit Soleil
Avatar

Send message
Joined: 17 Feb 03
Posts: 1497
Credit: 70,934
RAC: 0
Canada
Message 6421 - Posted: 11 Jul 2004, 8:22:15 UTC - in response to Message 6406.  

> Big caches aren't very useful for some projects.
>
>
> Preditor seems to run a pretty tight schedule and needs the results fairly
> fast.
>
>
> set your cache for 2 weeks worth of seti if you want, but your still only
> going to get as many as seti has or wants you to have at once time.
>
>
> It's fairly obvious that with a shortage of wu's people should go ahead and
> join another projects, even if they don't really want to.
>
>
> Expect wu shortages to be the norm rather then the exception, especially when
> seti@home classic shuts down. Unless they bring the other radio telescopes
> on line or do more extensive processing there is just too much cpu power for
> the work required.
>
>
> It's obvious that astro-pulse needs to start up ASAP. That will give a
> secondary project that many seti fans would find comfortable. And since it
> uses the same wu's from seti to reprocess including those that seti classic
> has already put into the finished and over with tape.
>

I agree 100% with you. Not enough recording for CPU power.

Here is a copy of a thread I made few days ago that seemed to anoyed a few
but thats the truth. period.

Are we still looking for ET ?

Looking at the server’s acquisition and splitter tapes status at http://setiathome.berkeley.edu/sstatus.html it looks like we are all crunching data that have been analysed many many times, I guess…

Is S@H still a project searching for ET signal or is it just for numbers cruncher’s competition ?
Isn’t it a waste of energy to “ stupidly ” crunch units for nothing else than numbers ? Should we use our processing power to some more “ useful ” humanitarians Projects like fighting AIDS ? At least until we can listen on other part of the sky, different frequencies, etc.

There are ten times more post in the number crunching section than SETI@home science. I was even planning to built myself a “ supercomputer ” dedicated to SETI, but what’s the point, really.

Or perhaps I am wrong.

Cheers
Marc from France




ID: 6421 · Report as offensive
STE\/E
Volunteer tester

Send message
Joined: 29 Mar 03
Posts: 1137
Credit: 5,334,063
RAC: 0
United States
Message 6431 - Posted: 11 Jul 2004, 9:21:55 UTC
Last modified: 11 Jul 2004, 11:40:45 UTC

Is S@H still a project searching for ET signal or is it just for numbers cruncher’s competition ?
Isn’t it a waste of energy to “ stupidly ” crunch units for nothing else than numbers ? Should we use our processing power to some more “ useful ” humanitarians Projects like fighting AIDS ? At least until we can listen on other part of the sky, different frequencies, etc.
==========

If you feel that way about Seti then why are you still here...???

ID: 6431 · Report as offensive
Profile Keck_Komputers
Volunteer tester
Avatar

Send message
Joined: 4 Jul 99
Posts: 1575
Credit: 4,152,111
RAC: 1
United States
Message 6535 - Posted: 11 Jul 2004, 15:01:27 UTC

We will need to recrunch some of the old data to be sure we are generating the proper results.

Even with new data sources we still need to keep using the old ones. The signals we are looking for may be transient, so even if we don't see anything for years we should still keep looking in the same places. Of course having more sources would be better and I hope the project will continue to try to get feeds from more and different telescopes.

John Keck
BOINCing since 2002/12/08
ID: 6535 · Report as offensive
Petit Soleil
Avatar

Send message
Joined: 17 Feb 03
Posts: 1497
Credit: 70,934
RAC: 0
Canada
Message 6543 - Posted: 11 Jul 2004, 15:25:24 UTC - in response to Message 6431.  

> Is S@H still a project searching for ET signal or is it just for numbers
> cruncher’s competition ?
> Isn’t it a waste of energy to “ stupidly ” crunch units for nothing else than
> numbers ? Should we use our processing power to some more “ useful ”
> humanitarians Projects like fighting AIDS ? At least until we can listen on
> other part of the sky, different frequencies, etc.
> ==========
>
> If you feel that way about Seti then why are you still here...???
>
>
Thats a good questions I have asked myself many times. Hopes I guess.
No, I believe in SETI's goals and all the sciences behind it. I just didn't understood why Berkeley was sending out WU that has been processed god knows how many times just to keep people in to the project and to keep crunchers happy. I think they should have been honnest with new users letting them know that they have more processing power then recording capabilities. Don't get me wrong when I say "stupidly crunching" I'm not saying crunchers are stupid !!! I fully understand the proud and pleasure of building your own computer farm and all the hard work making it run. I halso understand those who has it easy using 100's of corporate machines but just think of those who like myself are participating in the project using one or two machines. real Competitions for those is out of the questions. I am currently running SETI on my precious laptop and I'd rather be doind one real good and fresh WU per week than ten useless process. At least with boinc I know that when my machine is working hard it's doing it for science and not for a competition that I would never win in anyway. When boinc says no work available it means that there are no works available ! Some users have problem with that, I don't. I am working on a hugues cruise ship as chief radio officer and electronic engineer. I am in charge of all computers hardware so it would be easy for me to put a few dozens of machines to crunch. But whats the point ? It would not serve anything but having my name at the top and It would just waste precious energy for nothing. My ego would not rise and I would not be better then any others just because some stats says I have (or should I say some computers) did more processing job than 99% of total users. I will probably do it when it it will serve sciences. when astro pulse release or when we put more ear and recorder arround the globe. For those who forgot, SETI started as a screen saver. It wasn't consuming more energies than any other savers. How many machines arround the world are now running full blast 24 hours a day for nothing else than personal ego ? I could put some scientific calculations here to demonstrate how much energy we are wasting at the moment but just a few would understand. I have nothing against distributed computing or crunchers, I think it's a great global human project. I am just saying thet now, today, at the very moment, we could use this power for real science and not ego. Some might not agree with that and I fully respect them but it is my opinion and I am ready to defend it.

Is that answer your question ?

Best ragards
Marc

ID: 6543 · Report as offensive
STE\/E
Volunteer tester

Send message
Joined: 29 Mar 03
Posts: 1137
Credit: 5,334,063
RAC: 0
United States
Message 6544 - Posted: 11 Jul 2004, 15:47:43 UTC

Is that answer your question ?
==========

Pretty Much ... ;)

I bear a little animosity toward those that have access to corporate or place of employment Computers also. As far as I'm concerned they haven't done anything but get their name to the top with out the expense or time & toil it took to buy those computers like me. I only have 3 PC's but at least I've paid for them myself and I don't rely on somebody else's computers to get bragging right's within the Seti Crowd...
ID: 6544 · Report as offensive
1 · 2 · Next

Message boards : Number crunching : Keep enough work on disk to last between???


 
©2024 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.