1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

The Greatest Sticky in the known Universe

Discussion in 'TOSAH Archive' started by Feek, Jul 15, 2003.

Thread Status:
Not open for further replies.
  1. Feek

    Commissario

    Joined: Oct 16, 2002

    Posts: 28,234

    Location: In the radio shack

  2. Feek

    Commissario

    Joined: Oct 16, 2002

    Posts: 28,234

    Location: In the radio shack

    What is the SETI@HOME Project?

    Have you ever stared at the majesty of the heavens on a clear night? Have you wondered if, in all that vastness, with the billions upon billions of stars and untold numbers of planets, whether Earth is the only planet to support intelligent life?

    Not only is there a serious scientific study underway to try to answer that question, but you can be part of the effort.

    Seti@Home is a distributed computing project run by the University of California at Berkeley, and uses internet-connected computers around the world to analyse data collected from the giant Arecibo radio telescope in Puerto Rico.

    We all have fast computers, but is yours being put to good use, or does it just stay idle during times when you don't use it? All that power going to waste seems such a shame when you could be using it to run SETI@Home and making good use of your computer during idle moments. It would benefit Team OcUK if you were to join and show the world that the UK are the best overclockers in the world by climbing the World Rankings Ladder.

    SETI@Home involves searching the heavens for signs of extraterrestrial signals in the hope of finding intelligent life elswhere in the universe. The data is cut into managable chunks and distributed to home users for processing.

    All Seti@home actually does is a detailed statistical analysis of the signal received by the Arecibo telescope, looking for signals that cannot be attributed to background noise, interference or other naturally occurring signals. And if a signal is found that cannot be explained this way,then Mankind will KNOW it is not alone in the universe. And if that data unit is one you processed, you will receive credit for being part of the team that discovered intelligent alien life.

    If you want to take part, all you need to do is download a small program from Berkeley, and run it on your computer. It does not need to be installed as such, as it is a simple executable file so it doesn’t mess with your system settings. It connects to Berkeley (either automatically or at your specific command, depending on how you set your preferences), and downloads a small (about 350KB) data file for processing. Once complete, it uploads the results (about 7KB) and requests another data file.

    The idea behind this distributed computing project is that it doesn’t interfere with the normal operation of your computer. In fact, in can be set to only run when you are not using the computer for anything else and to stop running when you do use the computer – so it doesn’t interfere with your own computing. It merely uses the spare time when the machine would otherwise be doing nothing.

    In addition to the basic program, there are a range of separate tools and utilities available, with functions ranging from making it very easy to configure the program initially, to keeping a history of data units processed and even displaying a star map showing where your units were received from.

    Numbers have a certain fascination, especially in games, which SETI can be likened to. Each Work Unit released adds to your total and members become addicted to the Crunching, scrambling over each other in an effort to climb the Team ladder, often making amusing comments on the board as they do so.

    Within Team OcUK we have league tables and the team spirit is for members to try to "crunch" more Work Units than other members, rising through the leagues and "stomping" other members on the way. Of course, all such activity is fiercely defended by the potential stompee, and the net result adds to the team totals!

    In addition, teams compete against each other on the World Rankings Ladder and Team OcUK is currently in the top 10 of the top 200 teams worldwide, having competed against thousands of other teams. Our objective: To wave the UK flag for all the world to see by reaching the top position and proving to the rest of the world that UK overclockers are indeed a force to be reckoned with, and the finest team in the world. With your help we can keep the flag flying and hold our heads up with pride on an effort well deserved of praise by all envying nations. Honward and hupwards, the only way forward is hup is our battlecry, and the UK know better than any other nation how best to succeed. Give freely of your support and help Team OcUK and the UK achieve ultimate domination of the SETI universe.

    Whether you have 1 Work Unit or 10,000 it doesn't matter here at Team OcUK. Every WU counts, no matter how slow your machine may be, and contributes to our success. It's a team effort, and we all pull together for the sake of The Cause.

    The first time you start SETI@Home it will ask you a series of questions. Don't worry these are very benign and will not infringe on your privacy. The most important question it will ask you is for a valid email address. This is purely to identify you when you download units to process and upload completed units. They will not give out your address to ANY 3rd party and it is not published anywhere on the web site.

    One Important Note To Make Before Starting
    If you intend to run seti on your work computer, please ask your Boss/Network/Systems Admin before doing so. This will avoid any problems that might pop up. Make sure that you have that permission in writing. Under no circumstances try to gain access to any computers with out permission as this may endanger your job.

    Do I Need To Be Connected To The Internet Whilst SETI Is Processing It's Data?

    The only time you need to be connected to the internet is when you need to upload results and download new units to process.
    As the units are in the region of 350KB each it will take only a couple of minutes on a dailup connection to get fresh units to crunch, and to upload only takes a matter of seconds as the result files are on average about 6-10KB each
    ALL processing of the data can be done whilst offline

    Can I Run SETI@home On More Than One Machine Using The Same Account?

    Yes, We encourage this. Aslong as you enter the correct email address when setting up on a new machine then yes you can. There is no limit to how many machines you use for your SETI account.

    What Is A WU, Woo Or A Work Unit?

    SETI@Home divides all it's data into easy to process work units also called woo's or even wu's by members.
    Each work unit is a sample of a known location of the sky and a narrow band of frequencies at a known time.
    A work unit can take many hours to complete depending on the computer that is processing.

    Data will be recorded on high-density tapes at the Arecibo telescope in Puerto Rico, filling about one 35 Gbyte DLT tape per day. Because Arecibo does not have a high bandwidth Internet connection, the data tape must go by snail-mail to Berkeley. The data is then divided into 0.25 Mbyte chunks (which we call "work-units"). These are sent from the Seti@Home server over the Internet to people around the world to analyze. SETI@home looks at 2.5 MHz of data, centered at 1420 MHz. This is still too broad a spectrum to send to you for analysis, so we break this spectrum space up into 256 pieces, each 10 kHz wide (more like 9766 Hz, but we'll simplify the numbers to make calculations easier to see). This is done with a software program called the "splitter". These 10 kHz pieces are now more manageable in size. To record signals up to 10 KHz you have to record the bits at 20,000 bits per second (kbps). (This is called the Nyquist frequency.) We send you about 107 seconds of this 10 kHz (20kbps) data. 100 seconds times 20,000 bits equals 2,000,000 bits, or about 0.25 megabyte given that there are 8 bits per byte. Again, we call this 0.25 megabyte chunk a "work-unit." We also send you lots of additional info about the work-unit, so the total comes out to about 340 kbytes of data.

    And Dumping Would Be What?

    Many members refer to sending completed units back to Berkeley as "Dumping".

    To Stomp Is To What?

    When one stomps they gain a place in the league.

    What is Crunching?

    To crunch a woo is to process a work unit. Ergo crunching is the act of processing, computers which process work units are often called crunchers.

    What Is A Farm?

    The term Farm signifies a that a member has a number of computers that are running SETI@Home, from where this term originated is unknown but it has certainly stayed.

    Please read the below for information on how to get started.
     
  3. Feek

    Commissario

    Joined: Oct 16, 2002

    Posts: 28,234

    Location: In the radio shack

  4. Feek

    Commissario

    Joined: Oct 16, 2002

    Posts: 28,234

    Location: In the radio shack

    Setting Up the S@H Client (Advanced)

    To setup the Client software for SETI@Home please follow this guide

    First create a new folder on a location in your hard drive. It can be called anything you want really but it's much easier to keep it simple i.e. C:\Seti

    To start with you will need to download the CLI from here, Mirror 1, Mirror 2 and then place it in the folder you have just created.

    Then download Seti Driver from here - Seti Driver and place in the directory along with the CLI you have just downloaded.

    Create a shortcut to the Seti Driver.exe file in the SETI folder by right clicking your Start Menu, chosing "Open All Users" and navigating to the Startup folder.

    Run SetiDriver.exe Choose the following options:
    • Maximum Processes = 1 (unless you have a dual system then set it to 2, or a quad system set it to 4 etc.)
    • Desired Cache Size = 1
    • Hide Processing = Yes*
    • Auto Transmit = No*
    • Display Transmit=Yes*
    Now, click Save Config, and then Transmit (Do not click New Client). A black DOS window should pop up. After a few moments, it will prompt you to enter either "1" or "2".
    • Choose 1, if you need to set up a new SETI@Home account.
    • Choose 2 if you already have an account, just enter the email addy of that account.
    If you have chosen 1 to create a new account you will be asked the following question's in the DOS box which appear's when you attempt to download your first unit
    • Your Email Address:
      Enter a valid email address here - yourname@yourisp.com
    • Ok to show your email address on the SETI@Home website? (y/n):
      Really doesn't matter if you answer yes as it won't be shown anyway
    • Your name or nickname:
      This is the name that will appear on the SETI@Home statistic's page and the team's stats page
    • Ok to show your name on the SETI@Home website? (y/n):
      If you answer no then you will appear as a blank on all the stats pages, it is best to answer yes to this one
    • Your country number (Return to see country list);
      If you are from the UK then enter 213, if you are from elsewhere keep hitting return to find your country and then type in that number and hit enter
    • Your postal code:
      No idea why they ask this and it does not appear anywhere on the SETI@Home web site. The usual thing to do is just leave it blank.
    • Computer location: 0=unspecified 1=school 2=work 3=home :
      Just choose the one that best suit's your cruncher's location
    • You will now be presented with a list of your submitted information just choose yes or no to whether it is correct or not and the account is now activated and ready to use.
    The client will download the data it needs, the size of a unit varies between 341 or 351KB, and begin to process it. Once this occurs, return to the SETI Driver window.

    Choose the following options:
    • Maximum Processes = 1 (unless you have a dual system then set it to 2, or a quad system set it to 4 etc.)
    • Desired Cache Size = Number of workunits you wish downloaded at a time ( I find it best to download a weeks worth of units in case Berkeley has an outage or the server isn't working)
    • Hide Processing=Yes*
    • Auto Transmit=No*
    • Display Transmit=No*
    • Use SETI Driver=Yes*
    • Use SETI Spy=Yes*
    • Use 24Hr Clock=Yes*
    • Create SETILog.csv=Yes* (this creates a log you can open in
    • Microsoft Excel to view your progress)
    Click Save Config.

    When done the icon in the taskbar will be blue until a work unit has finished, when one has finished, it will turn yellow, at this point connect to the internet and press transmit before your cache of X work units runs out.

    *These are purely personal settings and have no impact on processing time whatsoever.

    For All Users:-
    Change your screen saver to "None".
    Navigate to Control Panel -> Power Options Properties. Change the power options for Hard Disk and System Standby to "Never".

    Certain Problems That Can Occur
    For Win 9X users:-
    Navigate to the C:\Windows\System directory and find the "Conagent.exe" file. Right-click it, and choose Properties.
    Navigate to the Misc tab, find "Warn if Still Active". Uncheck this option.
    This prevents some shutdown difficulties within win9x, though in some case's it doesn't have the desired effect and continues to give the annoying warning (I have found it shuts down with no problem in WinXP).

    To join Team OcUK go here

    When you attempt to join the OcUK Team, you will be propted for a password.

    To get this, please go to this link here and enter your e-mail address in the box provided.
    If you are running an ISP address or most webmail services are fine and you will receive the mail within 10 seconds or so. If you are running a mail address run by Hotmail or similar that has Junk Mail Filtering, check your Junk Mail Box if it doesn't arrive as it is most likely to be in here.

    Go back to the join page and enter your e-mail and password.

    If you have joined and only own one computer click here...
     
  5. Feek

    Commissario

    Joined: Oct 16, 2002

    Posts: 28,234

    Location: In the radio shack

  6. Feek

    Commissario

    Joined: Oct 16, 2002

    Posts: 28,234

    Location: In the radio shack

    OcUK SETI Stats

    Real time version of the Berkeley stats page:

    http://setiathome2.ssl.berkeley.edu/fcgi-bin/fcgi?cmd=team_lookup&name=OcUK+-+Overclockers+UK

    Personal Stats page (including links to your tiffys(certificates):

    http://setiathome2.ssl.berkeley.edu/fcgi-bin/fcgi?email=INSERT_YOUR_EMAIL_HERE&cmd=user_stats_new

    OcUK vs the World:

    http://gemini.setiatwork.com/cgi-bin/compare_teams.cgi?id=78617&type=0

    To view daily stats go here or here.

    Here is the latest version of SetiSpy which can give various information about the WU being crunched, team information and personal information:

    Direct Download
    Webpage

    Running SetiTEAM

    The following instructions detail how to set up a new copy of SetiTEAM, or how to repair an existing copy of SetiTEAM that has recently stopped working.

    1. If you don't already have SetiTEAM, download SetiTEAM 2.1i from here and save it to an empty directory.
    2. Start SetiTEAM.
    3. If you have never run SetiTEAM before, it should display a window asking for your internet settings. If it doesn't, right-click anywhere within the stats grid, and select 'Internet settings'.
    4. Enter your SETI user name. You can leave the email address blank if you wish.
    5. Change the SETI@home server setting to this:

      setiathome2.ssl.berkeley.edu/fcgi-bin/fcgi

      It is very important that you enter this setting exactly as shown, or you will not be able to download statistics. In particular, do not try to add http://. That will not work.
    6. When you're done, click OK to close the internet settings.
    7. Type in the team name (OcUK - Overclockers UK) in the box at the top left, and click Update.
    8. If all is well, SetiTEAM should download and display the OcUK stats.

    Daily stats are produced 99.9% of the time in the forum as well so look out for those too.
     
  7. Feek

    Commissario

    Joined: Oct 16, 2002

    Posts: 28,234

    Location: In the radio shack

    SETI Progress in Your Signature

    You can obtain a standard SETI workunit counter for your forum sig at one of the following sites (All the first list below are in German although relatively easy to work out.):

    http://www.wuschelkiste.de/setistat/
    http://seti.jesterweb.de/
    http://seti.firemans-page.de/
    http://www.seti-stats.de/setistats/index.php
    http://seti.familie-friedrich.de/

    We do however have a TeamOcUK English version hosted very kindly by Underboss Biffa. :) You can use your own images, and different types of text on the signature. Please check out the link below.

    http://setisig.3dgfx.net/setup.php
     
  8. Feek

    Commissario

    Joined: Oct 16, 2002

    Posts: 28,234

    Location: In the radio shack

    How to pick up your tiffy's...

    A lot of people seem to be asking this recently so heres how easy it is (although finding the tiffy is the biggest problem).....
    • Go here http://setiathome.ssl.berkeley.edu/individual.html
    • Enter your email address in the slot and press "Get Stats"
    • A list of tiffys will appear at the bottom of your user section. Click to view
    • If you want it in colour, try adding &color=1 or &color=2 to the end :)
    • Copy the URL and paste it here in a new thread for all to enjoy. Make sure it is placed in Tags.[/list]
      If you want a certificate with your current total and a cartoon in, enter the address in your browser and change to your E-Mail.

      [url]http://setiathome.ssl.berkeley.edu/cgi-bin/cgi?email=INSERT_YOUR_MAIL_HERE&cmd=print_cert&certnum=0&size=0&color=2[/url]
     
  9. Feek

    Commissario

    Joined: Oct 16, 2002

    Posts: 28,234

    Location: In the radio shack

    Jim's Guide to Improving Times
    Many thanks to TheyCallHimJim for his effort writing this guide.

    Thought I'd chip in my accumulated list of tips about how to make the most of SETI. Seems to be a whole lot of people popping up with pretty excessively long turnaround times, so I hope this is of help to some of yas...

    First and foremost, you MUST get the latest version (3.03) or your workunits will not be credited.

    Tips are listed in order of how much effect they have...
    1. First, ditch the GUI client. If you're in it for the sakes of sheer number crunching, then you've got no reason whatsoever to be using the GUI client. Go to the UNIX client page to download the appropriate command line version for your system. Under Windows NT, 2000 or 9x you want the version 1386-winnt-cmdline.exe , for Linux or other OSes, you've probably got enough know-how to figure out which version you need from the available ones anyway so you don't need my advice.
    2. Since the CLI version doesn't have the pretty graphics, you'll probably want to download SETI Spy. SETI Spy is a progress reporting tool that keeps track of how far the CLI version has gotten. It also does lots of other handy things like logging completed workunits, hiding the CLI client in the system tray, and showing you a sky map of where your current workunit is from, to name but a few. It's definitely not essential, but it makes using the CLI version a whole lot easier.
    3. If you're using a non-permanent Internet connection, you'll probably want to get hold of SETI Driver. SETI Driver is a utility that caches workunits so you don't have to be connected to the Internet constantly, meaning that you can eliminate that nasty time-wasting wait until you can get connected to download data. Even if you're on a cable modem or other permanent connection, it's still a handy little program, because it gives you a buffer in case of unexpected problems (server outages, ISP crapness, that kind of thing) and also allows you to transmit your completed data in the background so as not to waste a moment's processing time. SETI Driver also works perfectly well with SETI Spy, and I recommend using the both of them together, there are instructions at the SETI Spy site for how to do this.
    4. Assuming you've gotten the CLI version running OK, time for the real tweaking. First and foremost, leave it running all the time. Well duh! Stick SETI Spy and SETI Driver in your startup group and forget about it, you'll hardly know it's there. Just make sure that if you use SETI Driver as well that you make sure to send and receive a new bundle of workunits before you run out, otherwise it defeats the whole point of cacheing them.
    5. SETI@Home uses memory a lot; the vast majority of its work is shifting data to and from memory. So, make sure you're using the fastest RAM your system can handle, and if you can, crank your FSB speed up as well. Every little helps.
    6. A big, fast L2 cache also helps speed things up, so you can also help matters along by eliminating as much of the processor-level memory-resident stuff as possible from your system. Disable ALL screen savers on you system; if you must have your screen saved, set your system's power management to switch the monitor off after 10 minutes or whatever.[/list=1]

      The more of these steps you take, the faster you should be able to complete workunits. I've done all these things and even on my relatively slow system, my turnaround time has dropped from around 20 hours with the screen saver client, down to a shade under 10 hours on average. Even if you don't want to tinker with your system too much, steps 1-4 are pretty much essential if you want to make any real crunching progress. If you do nothing else out of the above steps, switch to the CLI client; it's the single biggest factor that can improve your turnaround times.
     
  10. Feek

    Commissario

    Joined: Oct 16, 2002

    Posts: 28,234

    Location: In the radio shack

    How to Build a Farm
    By Saracen

    If you don’t know what I mean by Seti Farms, I can almost hear the puzzled mutterings from here. So firstly, let me explain quite what I mean.

    A Seti farm is a collection of computers set up specifically to run Seti in a network environment. This is usually done in such a way that the hardware is carefully chosen to give the best available value for money for crunching Seti units, and with minimal expense on unnecessary fripperies – like case, keyboard, mouse, etc. An example can be seen in the OcUK Stomp Monster

    If you want to build something like this, there are a whole host of factors to be considered, including :-
    1. Safety – both physical and electrical
    2. Location – consider heat, noise and other environmental factors, like moisture levels in the air, especially in winter.
    3. Domestic harmony. Not a good idea to build this in the bedroom, as it is likely to irritate your partner.
    4. Physical design
    5. Budget
    6. Component Choice
    7. Remote administration
    8. Networking
    9. Remote boot [/list=1]


      Electrical and Physical Safety

      First of all, the caveat. What follows is my opinion, but I have no electrical or safety qualifications, so I am merely describing my approach to this issue. Neither I nor OcUK accept any responsibility for how you choose to design your system. This will highlight some points to consider, however.

      The power provision you need to make will depend somewhat on the size of your proposed farm. A system with 2 or 3 systems will make less demands on the power supply than 20 or 30.

      My approach to power was to run a separate ring main circuit specifically for the computer and radio gear in my office. I have a 30-amp fused connection from the main consumer unit running with heavy-duty cable inside protective trunking to a consumer unit in the office. From there, I have several separate ring mains running to different parts of the office, providing the ability to isolate computer bays and the radio gear. So I simply ran another ring main to the Seti Farm and provided enough wall sockets to power my Seti equipment without overloading any individual wall socket. It also means I can cut power to the entire Seti system at the flick of a single switch.

      I took this approach because I wanted plenty of facility to expand the system, and because the basic provision was easy to add : the office already had it’s own circuit and consumer unit anyway. But for powering a small number of machines, a simple 4-way extension should be adequate. After all, each machine will be running minimal extra hardware and make relatively nominal power demands.

      Most Seti Farms are built caseless, so the motherboard and electrical components are exposed. There are, however, no areas of exposed mains voltage. That only occurs inside the PC’s power supply, and that will be fully encased and protected. The motherboard and computer equipment run from 12v and 5v lines only, and the current is also very low. Sticking fingers in the wrong place may damage your equipment, but is unlikely to damage you.

      However, it is worth bearing two things in mind when deciding the physical design of the system and its location. Do you have kids or pets that may be tempted to touch the systems while operating? Is there any chance that the system may be knocked over – because it is likely to get expensive if you build a system like the one in the photo above, and it gets knocked off the utility room table.


      Location, Location, Location!

      How you build your system may well be determined by where you can put it.
      1. noise
      2. heat
      3. humidity
      4. loft/garage may have condensation problems. [/list=1]

        Mounting :-
        1. shelf – vertical stack
        2. shelf – side by side, and vertically stacked
        3. cupboard. If so, fans and cooling, doors, lock on door? [/list=1]
     
  11. Feek

    Commissario

    Joined: Oct 16, 2002

    Posts: 28,234

    Location: In the radio shack

    SetiQueue Setup Guide & Stashing Problems
    By Saracen

    SetiQueue is a proxy server designed to be used when running multiple Seti machines on a network, but there are benefits to using it with a single machine as well.

    SetiQueue is pretty easy to install and configure, but there is one point which catches most people out. When you first install SetiQueue, especially if adding it to an existing Seti and SetiDriver configuration, it appears to sit there and do nothing.

    In fact, it will do exactly nothing, until a Seti client attempts to connect through it. This initialises SetiQueue and from then on, it is plain sailing. You just have to remember to force the initialisation. I’ll come back to this later, but I stress the point now as it causes a lot of frustration when people see it apparently doing nothing.

    I have tried to go through the installation process step at a time. Which means the guide below is quite wordy. This is because so many people get tripped up, and need clarification. The process is not actually as daunting as the length of the section below would imply. Please stick with it. It takes about 5 minutes


    How to install SetiQueue
    1. Download SetiQueue

    2. Unzip SetiQueue into a Temporary folder of your choice (which will be deleted after installation), e.g. C:\TempSQ

    3. Run the install Program. This will ask where you want the program installed, and will put a shortcut into your Startup Group so that SetiQueue starts when you start the machine.

    4. Start SetiQueue

    5. When the SkyMap appears, go to \Settings\General and enter :-
    • Email address – the address you used to create your Seti account at Berkeley
    • Name, the ‘nickname’ you crunch under
    • Minimum Queue Depth (days), at least 1, but I suggest 7
    • Maximum Queue Depth (days), I set this to 10.

    6. Go to \Settings\Queue Server and set :-
    • IP Address, set this to the IP address of the machine you installed SetiQueue on. This will often be an address like 192.168.0.1.
    • To find the IP address if you don’t know it, under Windows 2000/XP open a DOS command Prompt and type IPCONFIG. Under Windows 98/ME, Go to the Start menu, select Run and Type WINIPCFG. If you have some sort of Internet Connection Sharing software running, make sure the Network adapter is selected in WINIPCFG and use the IP address shown.
    • If the machine is a standalone machine, not connected to a network, enter the loopback address, 127.0.0.1
    • Subnet Mask – set to 255.255.255.0
    • Port Number – Unless you have a reason to override the default (like you are upgrading from an old version that used 5512) then set this to 5517. This is the default anyway.
    • Choose the Connection Type. Default WinInet is the default and will usually serve. I use NTL here in the UK and have had better results by overriding the default NTL proxy server. See the NTL section in the FAQ for details.
    • Check ‘Override time for passthrough operations’ and ‘Override time for empty queues’. SetiQueue can be set to connect manually, when you want it to, or to connect automatically during times you define. There are certain operations, including initial setup and initialising new client types, that require these checked. However, once SetiQueue is running, these must be unchecked if you want to be sure SetiQueue only connects when you tell it to. If these are left checked, SetiQueue will connect when any of the necessary conditions are met, even if you have set Manual Connect only.
    • Connect Times are largely a matter of personal choice and the type of connection you have to the internet. Diallup users may wish to connect manually only to ensure the system doesn’t connect and rack up call charges without your knowledge. If this is the case, note the comments above about passthrough operations.

      Broadband users will probably be happy leaving the system to connect whenever it wants. However, at the time of writing, the bandwidth problems at Berkeley can make permanent connections irrelevant.

      The best time to connect seems to be while the Berkeley campus is not in full swing, and most of the US is asleep. This means weekends and, if you have to, between 7 AM and 12 Midday GMT (11PM – 4 AM PST). I limit this to 4 AM PST because it is 7 AM on the US East Coast by then and activity is starting to increase.
    • Report Stats to SetiQueue server returns some summary statistics to the SetiQueue website on units processed via SetiQueue. There is no real need to prevent these summary statistic being returned, but you can disable it if you wish.

    7. Forcing SetiQueue to initialise.

    SetiQueue can be used with the screensaver version of Seti or the command line client. The screensaver version, however, is so much slower than the command line client that any team member using it will be charged and convicted of treason, and sentenced to a term in the dunjinns at the tender mercy of Mr MindYerBeak. This term will last until you concede and switch to the command line client.

    For the recidivists among you, however, you can add the command line switch -proxy 192.168.0.1:5517 to the screensaver shortcut. Amend the IP address and Port Number as and if necessary.

    This is also the most frustrating part, especially while Berkeley are having their bandwidth problems, as getting a WU from Berkeley can take a while. The best times are mornings and weekends (GMT) as discussed above, but while testing the methods described here, I connected to Berkeley and downloaded a WU via the SetiDriver method at 5:00 PM GMT (9AM PST) successfully – if slowly.
    • Method 1 – command line client.

      Open a DOS box. Change to the directory in which you installed the seti client and run the client with the same switch as above, i.e. -proxy 192.168.0.1:5517

      For simplicities sake, rename the Seti client from xxxxxxxx to Seti and use a command line command like

      Seti -proxy 192.168.0.1:5517

      The IP address you need to use is the one for the machine on which you installed SetiQueue. Amend the Port number if you overrode the default.

    • Method 2 – SetiDriver

      On the main SetiQueue starmap Display, you will see the status line at the bottom showing ‘Get WU’ and, initially, ‘Receiving’. The ‘Receiving’ will then be replaced by a blue progress line as the WU downloads.

      If you click on \Browse\Queues menu option, you will see an entry for Queue 1 and a Red line while the unit is downloading. Once it is downloaded, and the queue is in operation, the line will change to mainly yellow, but different colour portions indicate the status of the Queue.

    8. Go to \Settings\HTTP settings and enter an Adminstrative password. With this set, you have access to SetiQueue options that you can’t access directly, and can also access and control SetiQueue from any machine on your network.

    9. Delete the Temporary folder you unzipped SetiQueue into (C:\TempSQ in this example) [/list=a]



    SetiQueue and Duplicate Units

    At the time of writing, Berkeley have been having bandwidth problems which result in the Seti project been capped and the servers often unavailable to receive uploaded results or to download new work units.

    The implications of this are twofold :-
    1. We strongly advise you use a cache and set the levels high enough to hold several days, maybe a week, of work units. SetiDriver is adequate for standalone machines but SetiQueue provides a much better solution for multiple machines on a LAN - and there is nothing to stop standalone machines using it either.
    2. The best times to upload or download are weekends and between 8AM and 12 midday weekdays UK time - when there are minimal numbers of students active at Berkeley and, during the week, when the bulk of the US is asleep.

    However, it has been noticed that SetiQueue has been classing some units as duplicates if they fail to get the correct response from Berkeley and removing them from the queue.

    Fortunately, they are stored in a subdirectory and can be returned to the queue. If you do this, it is very likely that they will be successfully transmitted (using the above guidelines on time) and credited to you.

    The procedure is :-
    1. Shut SetiQueue down
    2. In your SetiQueue directory, look for a subdirectory called \clients\duplicates. If it exists and has files in it, you have pairs of results files consisting of the results themselves and an INI file.
    3. Load the INI file into an editor like notepad
    4. Change the values in the last two lines to zero, so that they read

      Result/LastError=0
      Result/ErrorCount=0

      instead of something like

      Result/LastError=-66
      Result/ErrorCount=1
    5. Move the pairs of files up a directory into the \Client directory of SetiQueue
    6. Restart SetiQueue[/list=1]

      Providing you have set SetiQueue to connect at quiet times, as detailed above, you will most likely find that all the 'duplicate' units will successfully upload. However, it is possible you will find some genuinely are duplicates and will be rejected again.

      My results were that out of about 120 pairs, 15 were genuine duplicates and were consistently rejected, while the rest uploaded and were credited.

      You may also find a \client\dropped directory. If you do, apparently the same procedure works, but I have not experienced this.

      SetiQueue & Large Stash Problems

      Some people have had problems recently with Seti Queue and large stashes of results, when you try to transmit the queued results, instead of sending them it starts to delete them, with the message removing dangling result ini file.

      Fear not, there is a solution to this problem.

      Make sure you have a backup of all results before sending, shut down Seti Queue, move all of the results apart from 50 out of the client folder into a new folder; restart Seti Queue and send the 50 queued results. After these have been sent close down the queue again and copy 50 results out of your new folder into the client folder, restart the queue and send this batch.

      Repeat the above procedure as many times as needed to safely send your results.

      Also, because you have backed up your results before sending you have the advantage of having copies of any it mistakenly deletes, these can be copied back into the queue again to resend.

      I hope this helps people solve at least one of their Seti Queue problems.
     
  12. Feek

    Commissario

    Joined: Oct 16, 2002

    Posts: 28,234

    Location: In the radio shack

    SETI Remote Manager for Controlling Multiple Machines

    Here is a direct link to the zip

    Here is what it says in the readme.

    SETI@home Remote Manager v0.666
    (c)Demon 2001
    Proud Member of the LittleWhiteDog SETI@home team.


    Are you a Windows NT/2K administrator?
    Do you have hundreds, perhaps thousands, of NT/2K machines at your disposal?
    Do you want to run SETI@home on all of them?
    Are you too bloody lazy to install it manually on each of them?

    If you can answer yes to those questions then this is for you.

    What you need:

    1) This package

    2) The following Windows NT Resource Kit tools, preferably from Supplement 4:

    REGINI.EXE
    SC.EXE
    SRVANY.EXE

    3) The SETI@home command-line client. You *must* rename it to SETIclient.exe

    4) The file USER_INFO.SAH from an existing SETI@home installation. This file
    contains your SETI@home user details. If you don't have it then you can easily
    create it by running the SETI@home command-line client manually and supplying
    your details when asked.

    5) The names of each PC you want to install SETI@home on.

    6) Perhaps most important, you need to be using an account which has administrative
    privileges (either directly or through group membership) on the target PCs.
    I could re-write this so that you could specify specific accounts for each
    machine, but I'm too lazy to do that. Besides, I've got automatic admin rights
    on all of the boxes in my care; why haven't you?

    Put all of the files and programs into the same directory.

    I assume you're raring to go. Here is the information which is displayed if you run
    SR.BAT without any parameters. Oh yeah, nearly forgot: if you try to run SR.BAT from
    Windows it will run through without stopping and you won't see anything. You need
    to be using a command prompt. Scary, huh?

    =============================

    Usage: SR [ INSTALL | REMOVE | START | STOP | STATUS | HELP ] <computer name>

    <computer name> is optional. If you do not specify it then you
    must have a list of computer names in the file PCNAMES.TXT. In this
    file the computer names must each be on a separate line. The file
    must be placed in the directory

    E:\SETI\

    INSTALL

    This option deploys the SETI@home command-line client, configures
    and starts the service.

    Example: SR INSTALL

    This will set up the SETI@home service on all PCs
    listed in PCNAMES.TXT

    SR INSTALL HOMER

    This will set up the SETI@home service only on the
    PC called HOMER

    You must edit the file SETIREG.TXT if your network uses a proxy
    server (including SETIqueue, if you use it). You must put

    -proxy <address>:<port>

    at the end of the line

    Application = C:\SETI\SETIclient.exe

    For example:

    Application = C:\SETI\SETIclient.exe -proxy 192.168.0.1:8000

    REMOVE
    This option stops and removes all traces of the service. You will
    lose all progress made on work units currently being processed.

    START
    This option starts the service, if it is not already running.

    STOP
    This stops the service, if it is running.

    STATUS
    This reports the status of the service and current WU progress.

    If any errors are generated they will be in the file E:\SETI\RESULTS.TXT.

    =============================

    Every one of those options will take a computer name. I know I only gave an example
    for the first option (INSTALL) but it will work for the rest, too.

    If you want to manipulate one machine at a time then just put a computer name in the
    command. If you want to manipulate lots of machines then you need to create a text
    file with the names of the PCs you want to use. This file must be called PCNAMES.TXT
    and in that file every PC name must be on a different line, e.g.

    HOMER
    MAGGIE
    LISA
    BART
    MARGE
    FLANDERS
    SKINNER

    You get the picture. There must be no other text in that file. SR.BAT will assume that
    any text in that file is a PC name and will try to manipulate it accordingly.

    HINT: If you've got tons of PCs and don't know what they're all called then the Resource
    Kit can help you there, too...

    If you're behind a proxy server then you'll need to edit the file SETIREG.TXT. In that
    file is the line

    Application = "C:\SETI@home\SETIclient.exe"

    You'll need to change that line to

    Application = "C:\SETI@home\SETIclient.exe -proxy <address>:<port>"

    where <address> is the IP address of your proxy server and <port> is the port needed.
    If you use SETI Queue you can set the address and port of the SETI Queue machine as
    the proxy server.

    The STATUS option reads the progress directly from the STATE.SAH file. The value
    you'll see is a number between 0 and 1, where 0 is no progress at all and 1 is a
    complete work unit. The Windows NT scripting language can only cope with integers which
    is why you won't see any fancy multiplication to turn that figure into a percentage.

    Limitations:
    1) You must use an account which has admin priviliges on the target machines. This won't
    work otherwise.
    2) Only one instance of the service can be created on each machine. Once I finally get my
    greasy mitts on a dual-processor machine I'll rewrite this to cope with multi-processor
    systems.
    3) Depending on your situation you might want to disguise the processing. C:\SETI@home is
    a bit obvious, after all. If you can work out the syntax of the script file you should
    be able to rewrite it to install the service somewhere more discreet.
    4) The SETI@home command-line client *must* be called SETIclient.exe.

    That's it, as far as I can tell. This works on my systems; I hope it works on yours.
     
  13. Feek

    Commissario

    Joined: Oct 16, 2002

    Posts: 28,234

    Location: In the radio shack

    Other Distributed Computing Projects

    Following the demise of the Other Distributed Computing forum, here are the details regarding all the other projects that OcUK have a team in. This post was originally composed by Nobby and I thank him for his time and effort.

    Genome @ Home Classic

    The Genome@home client is part of the Folding@home architecture. To run Genome@home, you must choose "Genome@home" in the Client Type section of the Advanced tab in the Folding@home Control Panel, the first time you run the client.

    When prompted for your Company ID Number - put in 1057916486 which is the ID for OcUK.

    Official OcUK Team Stats are available from Here
    with Official Team Rankings available Here

    Keep watch on your output - Genome Spy


    Folding @ Home v3 (a.k.a Stanford Alzheimer and Amyloidogenic Disease Research Program)

    Download the latest client - Here

    When prompted for your Company Name/Number - put in 10 which is the ID for OcUK.

    Official OcUK Team Stats are available from Here

    Official Team Rankings available Here

    SLRI Distributed Folding
    Download the latest client - version Here

    When registering your details on the website, just select OcUK from the drop-down Team box.

    Official OcUK Team Stats are available from Here

    Project RC5-72
    Download the latest clients Here

    To join the OcUK RC5-72 team, you will first need your stats password - for this you will need to run the client and wait until it flushes work units back to the distributed.net proxy network. When your individual stats are being published (could take up to a day), you should click on the "I cannot remember my password" link to have your password sent to you.

    Once you have this password, go to the OcUK Stats page and click on the "I want to join this team!" link underneath the table.

    More information needed? Check out the FAQ


    United Devices Cancer Research Project (Phase II) (a.k.a. UD-Ligandfit)
    Download the latest client - version Here

    To join the OcUK team, click on the "Join this Team" button on the OcUK Stats Page.

    Team Rankings are available from Here (we are about ~200 / 31,489).
    [NOTE: stats are only updated every 24 hours, you may not appear straight away]

    NOTE: Be sure to uncheck the default options in your Device Profile for running other projects - as they don't have to be non-profit ones !
    (United Devices, by default, sign you up for projects that are being run for companies that are using the results for a profit.)

    If you have any problems, go Here for more information.


    Climate Prediction
    Download the latest version of the CP client from here.

    Once you have the software installed, and your username and password has been mailed to you, then you need to login to the user pages, and select to join OcUK - Overclockers UK.

    The team stats can be viewed here, once you have logged in.

    MD5CRK

    This project aims to find collisions in outputs of the MD5 hash algorithm using distributed computing (or a "Chinese Lottery", as mentioned in the previous story). The project will be considered successful if major crypto vendors - and developers in general - move away from the MD5 standard. MD5CRK is a unique distributed attack in that it attacks one of the most actively used crypto-related algorithm and will provide loads of data for crypto research. Other related projects only seek to decrypt a string of characters specifically encrypted to be attacked.

    Project is expected to take 2 years to complete.

    Clients are available for Linux, OSX, Win32, SunOS and also as a Java applet that can be embedded in any web page.
     
  14. Garp

    Capodecina

    Joined: Oct 18, 2002

    Posts: 10,561

    Location: Seattle

    List of Distributed Computing projects


    • Science
    • SETI@home - By far the best project there is! Help search for little purple men by analysing radio signals.
    • Analytical Spectroscopy Research Group - A more manual version of SETI@Home. Download a unit to work on, process it with one of three tools then e-mail the results back in
    • Evolution@Home - Powers research into Evolution, currently studying Mullers Ratchet, a form of genetic decay.
    • eOn - Helps calculate long time dynamics of systems, specifically the calculation of the time evolution of an atomic scale system, for use in theoretical chemistry, condensed matter physics and materials science.
    • Climate Prediction - Help to predict Earth's climate 50 years from now, using a large scale Monte Carlo system.
    • Distributed Particle Accelerator Design - Help design a more efficient particle accelerator. This project "simulates the pion-to-muon decay channel (grey cylinders surrounding a straight blue path) of the RAL Neutrino Factory front end design.
    • Lifemapper - This aims to create a massive predictive atlas of Earth's biological diversity.
    • Distributed Hardware Evolution Project - Help to design the next generation of self-diagnosing circuits.


      Life Sciences
    • Parabon Computation - Used for various medical research purposes. Runs as a secure java client.
    • Grid.org - Used for various medical research purposes, primarily working on Cancer in conjunction with the National Foundation for Cancer Research (NFCR)
    • Folding@Home - Simulate protein folding in an effort to better understand how proteins self-assemble or fold. Now also supports the Genome@home project.
    • FightAIDS@home - Helps design new drugs to fight AIDS.
    • Übero - Java-based client allows you to participate in for-pay projects, currently analysing genetic alignment algorithms.
    • Drug Design Optimization Lab - Help find oral drugs which can fight Anthrax, Smallpox, Ebola, and SARS deadly diseases for which there is currently no cure.
    • The Virtual Laboratory Project - Help screen molecules to design drugs to fight diseases, works by installing a grid computer client on your machine which schedules computations at their convenience. For permanent 'net connections only.
    • Distributed Folding - Much like folding at home, this simulates folding proteins to help scientists learn how proteins "fold and assemble into living cells."
    • Community TSC - Works to help find drugs to fight Tuberous Sclerosis Complex, a childhood disease that leads to benign tumors in multiple organs, including the brain, kidneys, heart, eyes and lungs.
    • Find-a-Drug - Help researchers find new drugs for major diseases at Find-a-Drug (non-profit.) Has successfully found some cancer-inhibiting drugs in the past!


      Cryptography
    • Distributed.net - Organizes projects like cracking (legally) data encryption schemes, such as RC5-72.
    • ECC2 - This is a distributed effort to solve Certicom's ECC2-109 challenge. Possibility of winning some money at this!
    • RSAttack576 - Is a distributed effort to solve the RSA 576-bit challenge, winner who finds the winning key will get $10,000 USD!


      Internet
    • Gómez Performance Network Peer - Evaluates the performance of large websites to find bottlenecks.
    • CycleTraders - Combines Distributed Computing with Peer-to-Peer computing concepts. Using the free CycleTraders client you can measure the response time of other users websites while they measure the response time of yours.
    • Grub - This is an open-source distributed internet crawler. It aims to produce the worlds most accurate, detailed search engine.
    • Capacity Calibration - Tests and monitors website performance in real-time with controlled capacity loads. Participants are paid $0.30(US) per hour of work.


      Financial
    • MoneyBee - Screensaver that uses neural networks to analyze stock market charts and indices to predict future trends of the indices. Users get summaries of reports produced for their own use in share trading.


      Mathematics
    • Great Internet Mersenne Prime Search
    • Proth Prime Search
    • *ECMNET - Finds new factors of Cunningham numbers.
    • n!+1 and n!-1 Prime Search
    • Minimal Equal Sums of Like Powers
    • GRISK - Search for K-optimal lattice rules.
    • MM61 Project - Search for factors of 2^(2^61-1)-1, a double Mersenne number.
    • 3x + 1 Problem - Finds 3x+1 class records.
    • Pi(x) Project - Calculates pi(x), for very large values of x.
    • Distributed Search for Fermat Number Divisors
    • PCP@Home - Looks for short cases of Post's Correspondence Problem.
    • Generalized Fermat Prime Search
    • Generalized Woodall Numbers
    • PSearch - Help search for the world's largest Proth prime number.
    • Seventeen or Bust - Help find the smallest Sierpinski number.
    • Factorizations of Cyclotomic Numbers
    • The dual Sierpinski problem search
    • ZetaGrid - Help verify Riemann's hypothesis. No one has been able to prove the hypothesis in 140 years. It is now considered one of the most important problems of modern mathematics. The project offers financial prizes!
    • Goldbach Conjecture Verification
    • pi(x) Table Project - Aims to construct a very large table of values of pi(x) for large values of x.
    • The Riesel Problem - Tries to prove that /k/=509203 is the smallest Riesel Number.
    • NFSNET - Uses the Number Field Sieve (NFS) to find the factors of increasingly large numbers.
    • Search for Multifactorial Primes
    • GMP-ECM factoring Lucas numbers
    • Distributed Emirp Project
    • Collatz Conjecture (Grid on Tap)


      Art
    • Electric Sheep - Uses a screensaver module to display an animated fractal flame while rendering frames for the next animation. Designed for Linux there is also a beta Windows client available.
    • Internet Movie Project - Render images for computer-animated movies in the Internet Movie Project. It is just for fun, still in development, but producing test images. Uses POV-Ray.


      Puzzles/Games
    • Genetic TSP - Uses genetic algorithms to solve a Travelling Salesman Problem, runs as a Java application in your web browser.
    • The Distributed Chess Project - Help train the next generation of computer chess players by helping this neural network evolve.
    • ChessBrain - Create a massive chess playing computer to rival Deep Blue, utilising idle processor cycles. Computers play chess against each other.


      Miscellaneous
    • DALiWorld - Not exactly what you'd call Distributed Computing. It's the world's biggest virtual saltwater aquarium. It populates it with fish, which currently just swim around. Occasionally they will swim off your screen and onto someone elses. Each fish carries a passport so you can see where its been, and who created it!
     
    Last edited: Sep 30, 2003
  15. Slackworth

    Capodecina

    Joined: Oct 17, 2002

    Posts: 10,013

    Location: Stoke

    Credit to the people who helped write this Sticky...

    Credit is due to all members of the OcUK Seti Team, but the following members were particularly helpful with the sticky....

    adwhitworth
    Berserker
    Biffa
    Feek
    Garp
    Kaiowas
    Markjcj
    Valor
    /\j C
    -<=[Tw18]=>-
    [ASSE]Arnie01
     
Thread Status:
Not open for further replies.