Posts by Brian Uitti

1) Message boards : News : SETI@home 8 released for Raspberry Pi. (Message 1786236)
Posted 9 May 2016 by Profile Brian Uitti
Post:
While I prefer to use the GUI BOINC Manager, for the Raspberry Pi Zero, I opted to use "Raspbian Lite" where these is no graphical frontend.

To get started with the command line tool: bloinccmd , you need to install boinc and boincmanager, as Claggy described with:
sudo apt-get install boinc-client boinc-manager


I refer to the link: https://boinc.berkeley.edu/wiki/Boinccmd_tool to work out the commandline details.

As stated, the general commandline form is
boinccmd [--host hostname] [--passwd passwd] command
.. For full permissions/access, under Raspbian/Linux you'll need to preceed the commmand with
sudo


With the commandline setup, it helps to have a pre-existing Account for each project that you want to setup. (although, there is a
--create_account 
option that you can use.)

"Be Brave" with the commandline tools. It's not very likely that you'll hurt anything! It's a pretty good way of understanding how it works.

Best Wishes with your projects.

// Brian
2) Message boards : News : SETI@home 8 released for Raspberry Pi. (Message 1777322)
Posted 8 Apr 2016 by Profile Brian Uitti
Post:
It should also work on other ARM linux machines that use the gnueabihf ABI.


Good Job! I'll be checking it out! Thanks!
3) Message boards : Number crunching : Raspberry Pi Support (Raspbian/Linux on Arm7) (Message 1775811)
Posted 2 Apr 2016 by Profile Brian Uitti
Post:

Is that normal on pi2?


Here are some of my runtimes from Project SETI@home/AstroPulse Beta

Raspberry Pi 2 Model B
https://setiweb.ssl.berkeley.edu/beta/results.php?hostid=78550
https://setiweb.ssl.berkeley.edu/beta/results.php?hostid=78181

Rasberry Pi 3 Model B (should be about 20% faster. Runs hotter than the 2B.)
https://setiweb.ssl.berkeley.edu/beta/results.php?hostid=78408

Raspberry Pi Zero (only one core)
https://setiweb.ssl.berkeley.edu/beta/results.php?hostid=78570

Keep in mind, on SETI@Home, the runtimes vary, depending on the data being processed.

// Brian
4) Message boards : Number crunching : Raspberry Pi Support (Raspbian/Linux on Arm7) (Message 1774675)
Posted 28 Mar 2016 by Profile Brian Uitti
Post:

Thanks Brian
I followed other link and managed build v8 version and ran on my pi2.
Because I use "make -j 4" for pi2 cpus, I got 4 tasks at same time and it will take 116 hours to complete one task. Is that normal on pi2?


Yeah, pretty much normal. You won't be breaking any records.
Here are my Raspberry Pi statistics: These are the projects that run "natively"... nothing to compile or install.



THe SETI workunits that I processed were from the boinc-seti package that I installed through the package manager (apt-get). Too many failures to bother with.
5) Message boards : Number crunching : Raspberry Pi Support (Raspbian/Linux on Arm7) (Message 1774557)
Posted 27 Mar 2016 by Profile Brian Uitti
Post:
Hi mimo
I just installed V7 version on my Pi2 and can't get any task to go.
Where can I get V8 on raspberry pi 2?

Thanks in advance.

Bowie

maybe some BETAs on SETi@home BETA site!

for now, only AP tasks come in v7!


Yes, SETI@Home/AstroPulse Beta has the V8 program that runs natively on the Raspberry Pi's BOINC Manager. I've been running on Pi2B and Pi3 for a couple weeks without error (Raspbian/Jessie).
http://setiweb.ssl.berkeley.edu/beta/
6) Message boards : Number crunching : Raspberry Pi Support (Raspbian/Linux on Arm7) (Message 1771443)
Posted 13 Mar 2016 by Profile Brian Uitti
Post:
Grant:

Don't get me wrong, I like the technical report that compares different systems.

I have a couple of thoughts:

1. some system has to take the bottom rung of the performance ladder.
2. sometimes running these projects mean that you have to use what you have available.
3. maybe we could suggest optimum projects for the weaker machines, like the raspberry pi

In the "SETI Classic" days, I had access to some "broken" laptop computers that we're one step away form the recycling center. These were mostly slow "Intel Celeron" based machines and had a range of defects such as cracked cases and broken screens. "Boot and Run" was a major "Goal Achieved" statement. In the end, they headed out for the recycling center, but there could be one or two that made a new life in a place where computers were just not happening.

Is it worth while? Of course it is. .. but if you're making a "purpose built" processing farm, there might be better choices.

I would encourage anyone to use whatever system they have to look into the concepts of the idea of sharing their resources with projects such as this.

Cheers.

// Brian
7) Message boards : Cafe SETI : Moon Hoax Theory (Message 884089)
Posted 10 Apr 2009 by Profile Brian Uitti
Post:
You should see the interview with Neil Armstrong.. where the interviewer asks for his reaction to "the moon landing was a hoax". There are people that will believe (dis-believe) anything.
8) Message boards : Cafe SETI : Moon Hoax Theory (Message 884086)
Posted 10 Apr 2009 by Profile Brian Uitti
Post:
There's one thing true about the movie "Capricorn One!" .. OJ can handle a knife.
9) Message boards : SETI@home Staff Blog : Donations to Help Keep Seti Alive (Message 582890)
Posted 7 Jun 2007 by Profile Brian Uitti
Post:
Jun 03 2007
Total Raised as of 31 May 2007 $181496.52 or roughly 34% of the $538000 goal.


Thanks for the update on the status of the fund raising effort and a bit about the spending direction. many outfits ask for money without a hint on what's going on.

// Brian
10) Message boards : Number crunching : Client Types web page broken (Message 575902)
Posted 26 May 2007 by Profile Brian Uitti
Post:
http://setiathome.berkeley.edu/client_types.php

Results in error:

"Warning: filemtime() [function.filemtime]: stat failed for /home/boincadm/projects/sah/log_galileo/connection_counts.out in /disks/setifiler1/home/boincadm/projects/sah/html/user/client_types.php on line 11
last updated: December 31 1969 16:00:00."

(I like that date!)

11) Message boards : Number crunching : Donation discussion (Message 165651)
Posted 10 Sep 2005 by Profile Brian Uitti
Post:
Not to mention the money that I've donated to the project by way of us government funding.


Actually, there's a good point.. the project owes it to it's supporters to make financial information available (at least in the form of annual reports).. especially since the days of the enron/worldcom debacle .. there's a new word going around (not often heard at the university) .. it's "accountability". ;-)


12) Message boards : Number crunching : Donation discussion (Message 165641)
Posted 10 Sep 2005 by Profile Brian Uitti
Post:
I'm intolerant of crybaby-ism.

My Star is bigger than yours! My self serving Avatar!


Maybe you should change your name to "Captain 'av-a-star" .. or "'av-a-bigger-star" LOL

13) Questions and Answers : Windows : A solution to "connection failed" (Message 155071)
Posted 22 Aug 2005 by Profile Brian Uitti
Post:
[quote
Hmmm, I'm suddenly having this same problem on one of my windows 2000 machines.. I've been using BOINC Manager 4.45 for some time now..
[/quote]

Interestingly enough, two subsequent reboots of this machine have successfully started the BOINC Manager without problems or need to utilize the "work around".

// Brian
14) Questions and Answers : Windows : A solution to "connection failed" (Message 154691)
Posted 21 Aug 2005 by Profile Brian Uitti
Post:
Okay, that worked for me, but isn't really a fix (more of a workaround).
How do I get back to BOINC starting properly? without having to run Boinc.exe from a command prompt?


Hmmm, I'm suddenly having this same problem on one of my windows 2000 machines.. I've been using BOINC Manager 4.45 for some time now..

YES. the work around get's it going.. but next time I reboot the machine (not too often on this machine).. this work around will have to be done..

I do have another machine that I shut down twice a day during the woek week.. (a laptop that travels to work) .. it would be real annoying if this problem shows up there.. (not yet!)

15) Message boards : Number crunching : Oh no 4.43... I cant keep up :) (Message 113018)
Posted 20 May 2005 by Profile Brian Uitti
Post:
Yep, it's a wild ride!! Got it going! (4.43)

// Brian
16) Message boards : Number crunching : Outages not good for business... (Message 104305)
Posted 27 Apr 2005 by Profile Brian Uitti
Post:
I've got no problem with down time.. I attach at least one other project to each machine that I run.. down time or out of work units.. no problem, my machines are busy with "something else".. all are served! (not a good idea to attach more than three projects on a machine! .. work units can run out of time!)

// Brian
17) Message boards : Number crunching : Question (Message 95147)
Posted 5 Apr 2005 by Profile Brian Uitti
Post:
> It'll take 9 1/2 hours of actual processing, but it'll only get one hour out
> of every ten -- so it'll take about 4 days if you run 24 hours/day.
>
> Ok, great! Just what I needed to know---thanks, Ned! I need to adjust my
> 'contact to server' by 1 day otherwise my Einstein units will be
> late---thanks again!
>

Obviously, turning in late is not a good idea.. but it's interesting to play with the numbers and see how you make out..

I'm currently running all of the boinc projects.. (SETI, Einstein, LHC, Climate and predictor) all set to "100" (..and Pirates at "10" on one machine only, LHC is currently out of work units.) .. but only setup 3 projects on each of my machines (I have 4 machines running) Also, I set my check-in at "1.25 days" .. this will give me more work units at a check in.. whole idea of the strategy there is, when a project goes down for a day.. I'll not likely run out of work units for that project.. if I do, then I've got the other projects cranking (like, I'm out of LHC work units right now.. but I've still got work to crank on.) If I want one project to be "favored".. like SETI .. I just have that one running on more machines than the other projects.. Still, the basic strategy is: Keep all machines cranking!!!

Meanwhile.. Have Fun!

// Brian
18) Message boards : Number crunching : Boinc V4.25 Stable Windows Version Release (Message 95145)
Posted 5 Apr 2005 by Profile Brian Uitti
Post:
>
> > >
> > > Column widths are preserved if you exit manually. - KR Jones
> > >
> >
> > Good point, I'll check that! Thanks,
> >
> > // Brian
> >
>
> Actually it was me who wrote that! :-) - Ertugrul Gokcen
>

oh! sorry about that.. Thanks. .. and "yes", it does as you say!

// Brian
19) Message boards : Number crunching : Boinc V4.25 Stable Windows Version Release (Message 95107)
Posted 5 Apr 2005 by Profile Brian Uitti
Post:
> >
> > With another strange lockup, I had to reboot again.. (The BOINC client
> was
> > locked up). sadly the tab column widths on all of the screens had to be
> > resized again.
> >
>
> Column widths are preserved if you exit manually. -KR Jones
>

Good point, I'll check that! Thanks,

// Brian
20) Message boards : Number crunching : Boinc V4.25 Stable Windows Version Release (Message 95106)
Posted 5 Apr 2005 by Profile Brian Uitti
Post:
>> I uninstalled the 4.19 BOINC client on one of my machines, and installed
>> BOINC Client 4.25 .. It's my Windows ME machine...> >
>> No real problems.. but when it first started up, it didn't show any
>> projects..
>> I attempted to attach a project, but nothing happened.. like with Windows
>> ME, REBOOTING is that natural solution.. and my existing projects showed up
>> and resumed (presumably) without a hitch.
>>
>> With another strange lockup, I had to reboot again.. (The BOINC client
>> was locked up).
>> // Brian
>
> Are you sure it's not just the benchmark running. I've seen it lock up Boinc
> for quite a while on my system. -Aurora Borealis

I kinda thought of that, but one would expect the Message tab would be showing that going on. .. and I did pull up the system monitor and the cpu was not pegged .. but memory was hammered.

// Brian


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.