Posts by popandbob

1) Message boards : News : Unexplained database slowness (Message 1931448)
Posted 22 Apr 2018 by Profile popandbob
Could be. Next we need to catch that process in the act of being enabled and then go look at the Haveland graphs for synchronicity.

Process currently running - only 3 Green Bank splitters running - remaining splitters are not running.
2) Message boards : News : Unexplained database slowness (Message 1931105)
Posted 20 Apr 2018 by Profile popandbob
i've just seen it , splitters speed limit ?

No, I think it is the process that knocks the splitters offline when the RTS buffer hits its max limit trip-point of around 620-640K. I've caught the splitters offline several times when the RTS buffer gets full. When the splitters go offline, it frees up resources for the file deleters and purgers to get caught up. You can see the behavior in the Haveland graphs.

The Splitters have that function build in. I suspect the function of this new process is to reduce the number of splitters running when under heavy db load to allow for other processes to take priority.
3) Message boards : News : Unexplained database slowness (Message 1927724)
Posted 1 Apr 2018 by Profile popandbob
The flaw still lies in the majority of the workload is in the workunit section. The forums/users add very little workload relatively.

If one was to separate the workunit database into separate sections (a lot of work no denying that) it would create the following situations:

Increased per CPU/GPU task numbers (we were told they were limited because of DB size)
Reduced weekly downtime ( instead of dealing with 1 large db you deal with only 2 smaller ones as only 2 would be fragmented, the others would be linear in function)
Possibility of still issuing work during weekly downtime (if ready to send were on a separate DB it could be done)
Reduced DB load (Instead of querying the full db to find tasks/wu's to delete/purge it could be done by a simple return first/last 20 results of my suggested waiting for purge db)

Dealing with a single large DB is only going to get worse as hardware continues to get better and more users join. It won't get any smaller.
4) Message boards : News : Unexplained database slowness (Message 1927626)
Posted 1 Apr 2018 by Profile popandbob
My 2cents worth...
I think perhaps its time for this project to rethink the 1 master, 1 slave database setup.
I do not see any reason preventing a multi-master database or several databases.
I think the best move would be splitting the database into several smaller databases. 1 ready to send, 1 in-progress, 1 waiting-validation, 1 awaiting purge.

For those wondering about finances/fundraising... ever look how many times this project has managed to raise the full value of its budget? I haven't seen it happen once in the 12+ years I've been here. The simple fact that this project is still here after falling that short on the budget continuously for that many years is a testament to the time and effort put forward by the crew here to keep the doors open.
5) Message boards : Number crunching : Panic Mode On (110) Server Problems? (Message 1920387)
Posted 22 Feb 2018 by Profile popandbob
From the info provided on the SSP down at the bottom of the page "(each file contains data for 14 channels)" but that doesn't answer your question.

I don't know the exact number of tasks per file other than guessing that 52.45GB รท 703KB per task = 74608 tasks. But how much overhead is involved with each task generation is unknown.

Don't forget the overlap and the file info adds a few bytes per wu...
6) Message boards : Number crunching : Intel and Gforce cards running together (Message 1904624)
Posted 3 Dec 2017 by Profile popandbob
Grant, I don't know if its just a SETI issue but when I was testing a new app over at folding@home the iGPU actually improved CPU output by 50%
7) Message boards : News : Breakthrough Listen data now flowing. (Message 1779177)
Posted 14 Apr 2016 by Profile popandbob
It looks like there isn't any astropulse for the breakthrough data? Is that due to the data itself or is there another reason?
8) Message boards : Number crunching : Panic Mode On (84) Server Problems? (Message 1373405)
Posted 30 May 2013 by Profile popandbob
Got 52 Tasks first try... All downloaded just fine.
Saw a funny message in BOINC a couple min ago though...

30/05/2013 3:52:34 AM | SETI@home | Sending scheduler request: To fetch work.
30/05/2013 3:52:34 AM | SETI@home | Not requesting tasks
30/05/2013 3:52:36 AM | SETI@home | Scheduler request completed

9) Message boards : Number crunching : Coming towards the end of an era in the top 100 machines (Message 1370185)
Posted 21 May 2013 by Profile popandbob

Not progress. Good Marketing. GTX295 versus GTX690.

The 690 is probably running 2 or 3 wu per gpu... judging by that computers RAC I would say it is.
10) Message boards : Technical News : Labyrinth of Light (Oct 06 2011) (Message 1159646)
Posted 7 Oct 2011 by Profile popandbob
Perhaps it would be a good idea to put the goodsearch/goodshop link on the donation page (since it wasn't mentioned in the email)
11) Message boards : SETI@home Science : SETI Brainstorm (Message 1156477)
Posted 27 Sep 2011 by Profile popandbob
I think that someone creative could turn this idea into something special...

What I'm thinking of is something along the lines of

Basically you make a program to visualize the data in 3D (in much smaller chunks of course!) and have users scan through it using a few built in tools but have the option available for users to write scripts to do what ever they want with the data (without being able to edit it)
It could be turned into a fun game I think... highest scoring signal wins!

12) Questions and Answers : GPU applications : Why does a GPU uses more FLOPS then a CPU, for the same job? (Message 984636)
Posted 28 Mar 2010 by Profile popandbob
I believe the reason behind it is the CPU can do some calculations in a single calculation which the GPU has to do in a couple.

It would be like asking the CPU to do 2x2 it would figure it out with one calculation of 2x2=4 where as the GPU would have to 2x2= 2 groups of 2 means 2+2=4 (2 calculations)
That's obviously not a true case but I'm sure you get the idea.

13) Message boards : Technical News : Lesson #1, Never use the word "fixed" (March 17, 2010) (Message 982676)
Posted 23 Mar 2010 by Profile popandbob
You've probably forgotten about this since its been over a year but you were going to do some sort of official announcement about the Good Search account you set up. Do you think you could get around to it sometime?
14) Message boards : Number crunching : EVGA has created a Dual cpu LGA1366 motherboard (Message 982334)
Posted 21 Mar 2010 by Profile popandbob
eVga are making their own power supply for that motherboard as well!
15) Message boards : Number crunching : New feature request...New Button (Message 982210)
Posted 21 Mar 2010 by Profile popandbob
It would also help other projects because you could tell boinc to not try and get GPU work from a project which doesn't have a GPU app.

16) Message boards : Number crunching : Petition For Seti@Home and Arecibo Funding (Message 950998)
Posted 30 Nov 2009 by Profile popandbob
The only emails that may be received are either those which you opted in for or they are related to the petition. You only receive other emails if you become a member.
If they do start to send other emails most email accounts have a method of blocking senders.
17) Message boards : Number crunching : Petition For Seti@Home and Arecibo Funding (Message 950818)
Posted 29 Nov 2009 by Profile popandbob
I would sign a petition, but I don't want to sign up for the service included with the signature, and according to the terms of service, there is no way to opt out.


What service is that? I don't think you have to sign up for anything if you sign.
18) Message boards : Number crunching : Petition For Seti@Home and Arecibo Funding (Message 950718)
Posted 29 Nov 2009 by Profile popandbob
Someone here on the forums wrote the letter, I cant remember who though... I was going to start this a couple years ago but never finished it until now.

19) Message boards : Number crunching : Green Star / Donation Controversy *THREAD LOCKED* (Message 948071)
Posted 18 Nov 2009 by Profile popandbob
For those who think its tough or they cant donate to Seti directly...

Select the University of California - Seti@home under charity

$0.01 per search is donated directly to Seti@home.

Eric was going to do an official post about this but never got around to doing it.

20) Message boards : Technical News : Back and Forth (Oct 28 2009) (Message 945390)
Posted 6 Nov 2009 by Profile popandbob
The latest version of MS FSX doesn't work virtualized, but I wonder if an older version would work?

X-plane has a linux and mac version... I dont think its as good as FSX but its better than older MSFS versions.

Next 20

©2018 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.