Posts by MRAO

1) Message boards : Number crunching : Always running older SETI application (Message 1999296)
Posted 23 Jun 2019 by Profile MRAO
Post:
A more interesting discussion than I thought it might be! I have unhidden the computer, at least for the moment, the one in question is 8001843. I do at (warm) times run it with temperature throttling in effect, I wonder vaguely if that might have some impact on this sort of thing. I could of course make it rerun the benchmarks, but it sounds like that won't help here. I'm reluctant to do a reset. However thankyou all for the splendid information.
2) Message boards : Number crunching : Always running older SETI application (Message 1999204)
Posted 22 Jun 2019 by Profile MRAO
Post:
Thankyou, the Gflops pointer is exactly what I needed to understand it. The 8.05 version is apparently faster on this PC for Seta main. Though. strangely, on BETA on the same PC 8.08 is recorded as faster (has a higher Gflops figure) than 8.05, which will be why my Beta tasks run that. Another puzzle.
3) Message boards : Number crunching : Always running older SETI application (Message 1999175)
Posted 22 Jun 2019 by Profile MRAO
Post:
Apologies if this has been asked before, but if so I can't find the thread. I notice that all my SETI CPU tasks run under the Windows 8.05 application, although I do have the 8.08 application downloaded and available. I am sure that when 8.08 was first released for SETI I used to have my SETI tasks running under that. I am also connected to Beta and the occasional task I get from that always runs under 8.08 and produces a good result, so there doesn't seem any inherent reason why SETI doesn't use it as well. Anything I can do to make SETI switch to 8.08 running, or is that a bad idea? Thanks.
4) Message boards : Number crunching : KWSN Windows optimized science apps - Share your results and problems! (Message 364871)
Posted 12 Jul 2006 by Profile MRAO
Post:
The quote may be wrong - nobody's perfect. It's how I remembered it :o)

Regards,
Simon.

It's how we all think of it now certainly! Try "Angnor" on Google, hardly anything but Chickens to be seen. ;-)

Thanks for your applications, a lot of work I'm sure for a result much appreciated by us all. Got my first Chicken WUs (sounds like a Chinese dish) and all validating ok. Mike
5) Message boards : Number crunching : KWSN Windows optimized science apps - Share your results and problems! (Message 364848)
Posted 12 Jul 2006 by Profile MRAO
Post:
Lol.

Incidentally, the lines about Robin include "who personally wet himself at the Battle with the Chicken of Angnor".

So there! Any errors in spelling were introduced by me, to befuddle you, the would-be investigator. Ni!

Regards,
Simon.

Well, I'm befuddled. I thought it was the Dragon of Angnor and the Chicken of Bristol, and best not to discuss the Battle of Badon Hill. But maybe it was different in other versions. I guess even dragons start small & cuddly. Mike
6) Message boards : Number crunching : Why are people really leaving? (Message 325290)
Posted 3 Jun 2006 by Profile MRAO
Post:

Are there any similar charts for computers - it would be more worrying still if the users leaving are the multi-computer ones. Mike


Here are the hosts-stats:

<img src="http://www.boincstats.com/stats/project_graph4.php?pr=sah&table=hosts">

Martin, thanks for that. The current number if I calculate it correctly is indeed dropping even faster than the number of users in RandyC's users chart (about 20500 hosts from the peak to the present). And if the growth in host numbers prior to April 18th is extrapolated upwards the drop since then is something that would very greatly worry me if I were a project Admin. Thanks again, Mike
7) Message boards : Number crunching : Why are people really leaving? (Message 325170)
Posted 3 Jun 2006 by Profile MRAO
Post:


This works out to: (188200 - 178000) / 188200 = 0.054198 (5.4198%) loss in active users. That's 10200 active users--A lot (especially if the trend continues), but not the end of the world.

Just trying to keep things in perspective.

Isn't the picture actually rather worse than that? Up to April 18th or so the number of users seems to be going up at a fairly steady rate, faster than it is now going down. There doesn't seem any particular reason why the number of new users a day would have slowed down much since then (or at least, I can't think of one). So the downward turn is probably more like double your figure or worse if that is taken into account.
Are there any similar charts for computers - it would be more worrying still if the users leaving are the multi-computer ones. Mike
8) Message boards : Number crunching : optimized Enhanced 5.12 app. (Message 323478)
Posted 2 Jun 2006 by Profile MRAO
Post:
My 2 pfennig's worth: this has to be the worst thing that has happened to SETI in a long time, even worse than mega-outages etc. I hope Crunch3R feels better about this soon, though whether he ever returns to SETI may be another matter.

Thankyou for everything you have done Crunch3R, and hopefully for that which is still to come. Mike
9) Message boards : Number crunching : Einstein@Home off the air. (Message 320765)
Posted 30 May 2006 by Profile MRAO
Post:
Bog Guy,

You're beating me!!! My Einstein box says only 131 hours and 12 minutes.

I've got one currently at 167 hrs & 36 minutes. I don't think you can get above 168 hrs which is a week. But perhaps someone can do better ...

Anyone know how to knock these back to zero when the project is back and settled? There's a trick with the clientstate.xml file but is there anything superior and safer?? Mike
10) Message boards : Number crunching : No new work from server since upgrade to 5.4.9 (Message 319557)
Posted 28 May 2006 by Profile MRAO
Post:

Personally, I wish he hadn't done that, but it's his call, not mine.

I do wish he'd include the app_info.xml file in the Optimized Application download.

I do notice one side-effect of this, which is that it confuses either the BOINC client or the application (I must read the code one day) into returning the highest numbered version_num in the app_info.xml file as part of the WU results as the application version that crunched the WU. So Crunch3R's current file is presumably causing the client/application to report that the current WU's were crunched with application version 519, not 512 or 515. See the last line of results reported using Crunch3R's very good application. This doesn't seem to cause any problems at present, but I guess only the Project Administrators could tell us if it might be an issue in the future.

And yes, it would probably help people new to Optimised Applications considerably if the app_info.xml file was packaged with the application, I agree. Forget to install it along with the optimised application and BOINC just seems to put the standard application back in place again.
11) Message boards : Number crunching : No new work from server since upgrade to 5.4.9 (Message 319531)
Posted 28 May 2006 by Profile MRAO
Post:
And since there are no optimized clients from Crunch3r for the 5.15's this won't help, as it's 5.15's they send out now.


Crunch3R's 5.12 application is compatible with the current standard SETI 5.15 application. The Seti Enhanced WU's that the server is sending do not have a specific version number, but the server does seem to get the BOINC client to check that you have a 5.15 compatible application or better. This is specified in the app_info.xml file by the presence of a <version_num parameter line in the xml.
I believe that Crunch3R's site is now delivering an app_info.xml file that claims compatibility with all versions from 5.12 up to 5.19 (clearly Crunch3R has some fortune teller capability) so if you have this version of app_info.xml you should be ok for the next few application versions, but would need to modify this file if version 5.20 or higher of the standard application appears. You need to keep an eye on this forum to see if that does happen in a few weeks or months. Mike
12) Message boards : Number crunching : SPARC? (Message 317966)
Posted 26 May 2006 by Profile MRAO
Post:
Einstein and SZTAKI has also applications for Solaris on SPARC.


That's a thought. My box would be a dedicated cruncher. I was just worried that it would suck up a lot of power and not crunch much.


Or of course calculate how much you are saving each week by not running it, and put that towards a new Intel low power D (for "Death Star") core processor machine, capable I believe of vaporising a small snowball over the period of about half an hour. Mike
13) Message boards : Number crunching : No New Work - why ???? (Message 311357)
Posted 20 May 2006 by Profile MRAO
Post:
Or get the latest app_info.xml file from Crunch3r's site.

If he does that, he also needs to get the new optimized app. He only crunched 4.18 before, so just the app file will not help.

Plus now with Enhanced 5.15, I am unsure where that is in the works.

The app_info.xml file on Crunch3r's site still appears to be back at WU version 512, so if downloaded it needs the <version_num> line about 10 lines from the end changing from 512 to 515 to work with current WUs (and updating in line with future WU versions from then on). The other references to 512 in that section relate to Crunchr's application on the same site, which is still 512 but quite happy crunching 515 WUs if app_info.xml is adjusted, or so it seems.

See comments from Josef Segur on this topic in other threads for detailed information about this and confusion with Beta WUs.
14) Message boards : Number crunching : optimized Enhanced 5.12 app. (Message 307944)
Posted 16 May 2006 by Profile MRAO
Post:
I got a "error 403: Forbidden!"
when trying to reach Crunch3rs site: http://www.guntec.de/Crunch3r

Is this something temporary ?


try http://calbe.dw70.de/index.html, maybe he has moved. Mike
15) Message boards : Number crunching : CPU overheating (Message 307921)
Posted 16 May 2006 by Profile MRAO
Post:
my cpu on my laptop overheats and shutsoff while im sleeping... is there a way to tell BOINC to use say.... only 80% CPU? there was a way to do it in folding@home's application but i cant seem to find it in BOINC.

Look for threadmaster at http://setiathome.berkeley.edu/download_network.php - does the job nicely for me. Mike
16) Message boards : Number crunching : Computation errors on Seti Enhanced (Message 303408)
Posted 12 May 2006 by Profile MRAO
Post:

5/10/2006 1:02:25 AM|SETI@home|State file error: no application version setiathome_enhanced 513

There is an SE 5.13 application, but we're testing that in Seti Beta. Sounds like there was a cross-over going on here. Or is it released here as well?

I've now got a batch of errors about 514 too:

2006-05-12 15:57:55 [SETI@home] Sending scheduler request to http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi
2006-05-12 15:57:55 [SETI@home] Reason: To fetch work
2006-05-12 15:57:55 [SETI@home] Requesting 1842 seconds of new work
2006-05-12 15:58:00 [SETI@home] Scheduler request to http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi succeeded
2006-05-12 15:58:00 [SETI@home] WU refers to nonexistent app_version: setiathome_enhanced 514
2006-05-12 15:58:00 [SETI@home] Can't link workunit 30mr99aa.29736.4305.1034646.3.91 in sched reply
2006-05-12 15:58:00 [SETI@home] WU refers to nonexistent app_version: setiathome_enhanced 514
2006-05-12 15:58:00 [SETI@home] Can't link workunit 30mr99aa.29736.4305.1034646.3.98 in sched reply
2006-05-12 15:58:00 [SETI@home] link_result: nonexistent WU 30mr99aa.29736.4305.1034646.3.91
2006-05-12 15:58:00 [SETI@home] Can't link result 30mr99aa.29736.4305.1034646.3.91_2 in sched reply
2006-05-12 15:58:00 [SETI@home] link_result: nonexistent WU 30mr99aa.29736.4305.1034646.3.98
2006-05-12 15:58:00 [SETI@home] Can't link result 30mr99aa.29736.4305.1034646.3.98_3 in sched reply
2006-05-12 15:58:00 [SETI@home] No work from project

any clues yet anybody what is causing these?
17) Message boards : Number crunching : uninstalling boinc and seti (Message 302064)
Posted 10 May 2006 by Profile MRAO
Post:
I got a program called 'wipe.exe' that does a good job, lol.

Deleting command.com or ntldr.com (depending on Windows version) is almost as exciting....

Absolutely. My boss came in one day complaining his laptop wouldn't work any more. I asked him if he had done anything. No he said, just deleted some junky looking files in the root of his C: drive. Hmmm.

But you can get it back working again then, just put the drive in another similar PC and copy the missing junky looking files back again! Wiping it completely is more permanent. Mike
18) Message boards : Number crunching : ut oh, all the threads are empty (Message 301182)
Posted 9 May 2006 by Profile MRAO
Post:
Last time 6 days were missing initially, this time any post in the last 14 days are GONE. 15 day old posts and earlier are still there.

Ah, Groundhog fortnight is it?
19) Message boards : Number crunching : ut oh, all the threads are empty (Message 301171)
Posted 9 May 2006 by Profile MRAO
Post:
I like it. Makes it easy to moderate. :)

Let's start all over with a clean database. :)


I think searching for an answer at the moment is going to have rather less than 95% success! Mike
20) Message boards : Number crunching : CPU-usage (Message 298584)
Posted 7 May 2006 by Profile MRAO
Post:
Thank you. But it doesn't match.


What is the name of the task that is running at 100% (see Task Manager)? The name of the SETI application is in the process of changing to the Enhanced application. I found I needed to add the new Seti application name to the Threadmaster parameters so that both the 4.18 and 5.12 applications are suitably throttled. If it will help you please post your seti.reg file here & I'll tell you what changes I had to make to the Applications key to slow down 5.12 too. Mike


Next 20


 
©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.