I got a late start, but my PC's been pecking away at a cure for Cancer for the last two days. woke up this morning and could see it did well last night! This is pretty cool.
Printable View
I got a late start, but my PC's been pecking away at a cure for Cancer for the last two days. woke up this morning and could see it did well last night! This is pretty cool.
Thanks for helping out.:)
How can I change where the data folder is located? I installed it and selected my D: for the data. That's a mechanical drive though, and I'd much rather have it on my C: (SSD raid0 array), so my mechanical drive can spin down to reduce power draw and decrease system noise. Thanks! :up:
Edit: Nevermind. Figured it out. Had to change a registry key ;)
Today I shut down my rig and pulled the HS off to put the stock cooler on until my water cooling arrives. I booted back up in OSX and installed the Bionic client and have been running in OSX since.
Does this mess anything up with my production? would windows be more productive than OSX?
Both OS are 64bit but I just installed the standard download from WCG so not sure if the client is 64bit or not.
Supposedly the 64bit client is more efficient on a 64bit OS.
Here is the download:
http://boinc.berkeley.edu/download_all.php
Thanks!Quote:
Originally Posted by retro77
Hey everyone! I've been away a little, but thanks for the answers to my earlier post.. And I've learned a lot reading last few pages..
But I do have few more questions:
- how did we do for the "WCG week"? Any stats and comparisons against weeks before Movieman sent invites?
- how long does "Result status" logs data? I know I had some errors before, but now when I filter, I have nothing. Seems like it only logs for one week, or there about..
- how come that I have 67 units "in progress", when I'm crunching with 3 computers (total of 12 threads). Acording to this:
.. "in progress" units are ones that are "currently crunching", so shouldn't I have just 12 units in the list? Or does BOINC client "pre-loads" a bunch of units to all of my computers? Dates of all units are from today, so they are not some old units or anything..Quote:
Originally Posted by Pointhore
And I'm glad to say that, at least for the period shown in result status, I've got no errors of any kind.. Just units in progress and waiting validation, and those that have been checked and validated just fine. I'm currently crunching just 3 projects: Cure Muscular Dystrophy (phase 2), Fight Childhood Cancer and Conquer Cancer. Seems that these are getting done just fine on my computers, so I'll leave it at that.
What else? Nothing! all peachy :D Cya all later!
EDIT: Oh right, and in 3-4 days I should be at 1 million points, that will be almost exactly 1 million in one month :D But I'm glad that I've relocated i7 server to office (few days ago).. let them care about the heat dissipation and electric bills :D
We did more than double our normal output for the challenge week. From roughly 7m per day to 17m per day.
Results status page only keep things there until they get off loaded form the WCG servers and sent to the scientists. This can be any where form a week to a month.
In Progress means WUs that have been downloaded to your machine but you have not finished crunching and sent back the results. Yes, WCG does front load WUs. You can control this by adjusting your cache settings on the device profile you have each machine set to.
http://www.worldcommunitygrid.org/ms...viewDevices.do
Thanks for the answers.. And great news about the increase of points generated.. nice.. And (dooh!) I've just realized I could have looked at it myself on the WCG site.. If anyone is as lazy as I am, here is the link that leads to WCG statistics of XS team for last 60 days: Link
going from 7 mil to 17 mil in just 25 days, nice.. I just hope that most people will stay in the team and keep crunching at least for a while :) Just a fun fact: if we stay over 17 mil points per day, we could reach #2 spot in team rankings by the end of year ;)
Question: what to do to see hosts at Free-DC stats??
Quick question, is BOINC included in WCG or are they two separate crunching programs? Should I be doing one over the other?
i'm pretty sure BOINC is just the "manager" for the various DC projects out there. You crunch through boinc if that makes sense. Correct me if i'm wrong anyone
Think of BOINC as the manager app.Quote:
Originally Posted by Utnorris
You install it and then "attach to project" and pick which one you want.
WCG is one of these.
Make better sense now?:D
beat ya :)
@ ggdh ... here is a link to your summary page, click around and have fun :up:
http://stats.free-dc.org/stats.php?p...e8e052bb6e1d76
@UTNorris ... BOINC is the framework that lots of DC projects run on, including WCG. If you download the WCG client it will autmatically install all of the necessary bits and pieces. If you are running a 64 bit OS then then you need to get the client from BOINC directly.
http://boinc.berkeley.edu/
This will also install all of the necessary bits and pieces.
Once you have this set up, you can join the XtremeSystems team by clicking the web address below and then click on the "Join Now" button on the page that appears.
http://www.worldcommunitygrid.org/te...mId=69B8364XP1
@ Snow Crash, i meant my own hosts, own machines. Like device manager in worldcommunitygrid.org. If at free-dc you will search for hypernova, movieman or any other "power user" you will see machines they are using. Not me. Actually not only me, but all "small fishes in the pond". Heaving a hosts stats is a privilege of big crunchers??
Host stats are not restricted to or for any special status ... seriously, sometimes that site just gets a little messed up (I think it might just be in between updates which can take quite a long time). It usually shows my three machines whe I look at my WCG specific stats but none are showing when I looked just now. :shrug: give it a little time and check again. Whil;e BOINCStats runs on a completely different cycle it is another good place to see stats. Hold on a second ... perhaps you have your hosts hidden?
@ggdh Sign in to WCG go to my profile and at the bottom you will see "The following information is used on websites which integrate with World Community Grid:" ....there check the "show hosts" box
I'm guessing that it takes a while before free-dc will notice that i changed the settings. But OldChap - your're a star:up:
yup, I only come out at night :)
Looking better now http://stats.free-dc.org/stats.php?p...cg&name=681266
Yeah, now everyone can see my lousy crunchers:cool:
i have a few WU's that came up inconclusive and was wondering if this is because my power went out?
i had the power go out once before and all my WU's were all valid,thats why im asking.is it like the lottery?win some lose some?
Nope, they will generally get credited in a couple days..Non issue..Quote:
Originally Posted by radaja
cool, thanks dave
How does the BOINC manager decide which WCG project to download? I have currently HCC, FAAH & CMD activated but as a FAAH and CMD take much longer , HCC is kind of being disadvantaged(is that proper english?).
Is there any way to manipulte this, as I like HCC the most but I i want to crunch the others, too, but just in a balanced way.
Thanks Soya
The only way i see for that, is if you have more than one pc, and use the profiles to distribute the projects per pc.Quote:
Originally Posted by soya_crack
You could use one pc at the home profile for HCC, one at work profile for another project, and another attached to school profile for another project.
Or you could crunch one project exclusively for a week and change the project each week.
Good morning my fellow crunchers :)
Is there an easy way to find out how many points a given processor usually produces?
I've got a couple of LGA 775 boards, one is currently using an E8400 CPU and averages about 1600 Boinc points a day. The other board is still in the box it came in, waiting for a CPU.
I want to buy two Socket 775 Quad CPU's for these boards so I can maximize my points and power usage. So CPU cost is a big factor for this upgrade, but I also want to get the most points for my $$.
In Canada I can currently find the Q8200=$174, Q8300=$159 , Q8400=$199. Obviously the Q8300 catches my eye as the cheapest of the bunch, but I don't know how many points it will produce.
Any help would be appreciated, and if you have any 775 Quads laying around to help our crunching cause, I would appreciate hearing from you (PM please as I'm probably breaking a rule asking). I've already looked in the for sale forum to no avail, but thought I would ask here as the WCG community seems to be more willing to help out their fellow crunchers.
Regards,
John
You can try read the PPD Predictor thread, as it might give you a good estimate on WCG PPD you might expect.
These Q8x00 are Yorkfields, so you just have to multiply 5.5 by the speed of the processor to get estimated average wcg PPD.
Obviously, overclocked processors wiil get more points, but if you are keeping them at stock, i would estimate this:
Q8200 (2333MHz) about 12830 wcg PPD (1830 boinc ppd)
Q8300 (2500MHz) about 13750 wcg PPD (1964 boinc ppd)
Q8400 (2660MHz) about 14630 wcg PPD (2090 boinc ppd)
Thank You very much for the info.Quote:
Originally Posted by pfm3136
Sounds like the $153 Q8300 would be the way to go as far as $/ppd. Add a little juice and that should add nicely to the point total :)
John
John
I came home this afternoon to find that boinc had frozen on my linux machine :( Restarting boinc didn't help (it freezes immediately after starting) and neither did restarting the machine. Is there any way to save the work or are my options just delete everything and start over?
Don't know if it's the same problem, but i did have to abort the beta work units.Quote:
Originally Posted by Red Maw
That's what I planned on doing but it looks up immediately after loading :(
You may need to reset the project. Also if it's overclocked you might need to drop it by a fraction. Maybe.
There's no OC on that machine but I have had problems with it in the past so that isn't out of the question. It appears that the last 6 WU returned ended up as errors :(
I reinstalled boinc but it won't let me attach to a project anymore ("attach to project" button/menu option just don't do anything) so I'm going to rip everything related to boinc out by hand this evening and try again
hi everyone i wanted to know if you can find out what rank i im in the team
becuse going to wcg it says 1962 members (14) sierrabound
im asuming were 14th in the world?
so is there any way if seeing your own personel rank (by team member)
if you know what i meen thanks :)
@rubber5:
Here you go:
http://boincstats.com/stats/user_gra...=wcg&id=681456
There is a line there for Position in Team. Good luck!
thanks for thatQuote:
Originally Posted by retro77
my god so much info how do you understand it lol
information overload lol:confused:
I started getting massive units on one of the machines (not the fastest though). Takes up to 27 hours to complete.
Any reason why? (no complains here:))
i think it becuse of the day's work load i think bionic downloads a 24hr periodQuote:
Originally Posted by ggdh
you have to go to
advanced then preferences then go to network usage
in the additional work buffer set it to 0.10 and it will reduse the ammount it will download for working treads if im right on this
im new to this aswell so not sure on everything
hope this is what you wanted to know
I was about to post the same thing , suddenly all my Fight Aids are talking at least 12 hours and Help Cure muscular dystrophy up to 9 hours , thought mu PC is getting slower in time :DQuote:
Originally Posted by ggdh
I doubt if it will change anything. It's set to 0.2 since 45 days and nothing like that happened before.Quote:
Originally Posted by rubber5
i have tried to put all the deatals on 0.00 and it still downloads loads of tasks does anyone know how to stop itQuote:
Originally Posted by ggdh
all i want to do is download tasks to do for that day i only run it for about 12hours in a day but only for 2-3 days at a time max and i dont want to have tasks getting done to late in case it lowers my stats in any way
thanks everyone :)
You can hit "No new tasks" in Projects tab. Try to set work buffer at 0.1. When i have it set to 0.2 on the machine that gets large WUs i don't have more than 4 units waiting in a queue.
Don't pay any attention to the "to completion" time estimates in BOINC, especially estimates on WU's that are waiting to run. They are only estimates and can be WAY, WAY off! I have found that if you run projects that have WU's with highly variable run times (such as HCMD2 that can run anywhere from a few minutes to 12 hours), it really wrecks havoc on BOINC's completion time estimates for all projects, ie- the estimates will swing wildly up or down.
As for your other problem of downloading too many WU's, I'm, not sure what is happening there. "Connect about every:" should be set to around 0.1 days and "Additional work buffer" can be set anywhere you like but 0.1 is a good number if you want a small queue. Maybe don't set them to zero as I don't know what effect that has. Be sure to hit "update" in the projects tab for the settings to take effect. If all else fails hit the "no new tasks" button so no more WU's are downloaded.
Few units are already done. Longest took 26h to complete so BOINC estimation is pretty accurate to me. Another example: 61% done, took it 8h43min, 6h28min to go.
I just crunch 'em all....
thanks for that everyone
i will give it a go and see what happens
i dont know if im right on this but if i cruch for 12hrs
im getting about 1600 points does this seem right foe an i7@3.5 ghz with ht on
here is my stats becuse i cant undertstand them at all lol
http://boincstats.com/stats/user_gra...=wcg&id=681456
I have a question about WCG points and validation time.
When I send in completed WU's late in the day they don't get validated in time for that day. Maybe they will more than a day later.
Do the send WU's count for the "Returned Results" total for that day?
And for what day do the points count? For the day that the WU's were send or for the day that the WU's got validated?
The points for each validation count to the day in which they validated, not when they were returned. hth
Why do some WU's run in "high priority"? As far as I can tell it only happens on one PC and with different projects. Why doesn't it finish the WU that it is working on, before starting a new one in high priority? Is that something I can change some where? (I couldn'd find anything)
No, you can't change it and you wouldn't want to anyway.
When a unit is promoted to "high priority" it's because it's gotten very close to it's due date, either because it's been in your cache for too long or more likely because WCG sent it out as a "repair" unit (which means someone else didn't get around to completing it in time). The unit that got suspended to make way will pick up after the urgent one is done.
Looking to install BOINC as a service, my bro is a :banana::banana::banana::banana::banana: and always clicks out of the icon.........
when a app/proj is finished & its status is 'ready to report' is there any way to hurry it along? aka get it to report already?! :p:
Click through Project tab > select the project from the list > click the "report" button on the left.
sweet, thanks.. now I can format & try to see if I can run R0 of the ICH5R instead of the SI :up:Quote:
Originally Posted by D_A
Can anyone help me configure WCG and GPUGrid on the same machine?
Phenom 2 965 @ 3.60GHz NB @ 2.4GB
2GB @ DDR1450 6-6-6-18 Ganged (I have my reasons for not running unganged)
GeForce GTX260 216/55nm @ stock
Questions include how much CPU usage should I give GPUGrid on their site? Do I need to change any WCG preferences?
It makes no difference what you set CPU usage preference on GPUGrid, it uses what it needs to support the GPU. If you are runing fermi then there is an environmental variable you can set to let it use a full CPU thread in conjunction with the GPU to get the best runtimes possible but again, this is not controlled by the CPU percentage ppreference on the website. None of this matters with 2xx based cards ... they only need/ use a very small perecnt of your CPU no matter what you set anything to.
how to I prevent WCG from using my main OS hdd? can I put a 2nd hdd in, install WCG to that, so that it uses only that 2nd hdd for all its needs?
Without having tried it on windows, I'd have to say: it'd have to be possible and should be quite easy.
My syistem is runing with ATI cards and bionic is working fine. When I install gtx470 and nvidia drivers (clean install) bionic wont run . Program is up but not crunching. Even when i reinstall it. Can somebody give my advice what to do because i tryed everything exept fresh win install.
Did you Tools>Attach to project/project manager?Quote:
Originally Posted by roki977
roki ... WCG does not use any GPU, have you tried attaching to GPUGrid?
GPUGrid is entirely different project but manyu of us crunch both WCG and GPUGrid ...
We have a subforum here on XS for GPUGrid ... stop by and ask away if you are interested :up:
I am not crunching with GPU . I had milkyway attached before with 5870 a maybe its because of it. Simply when i put gtx470 instead 5870 WCG wont run. BIONIC is runing but nothing is happening when i add WCG. It wont start crunching.
sry for my bad english.
I will try again and write exacty what happened but everytime i try am losing time . Thx .
There is something confusing with DC Stats: daily score is not equal to the sum of results of each of machines. For example yesterday it reported ("Last 28 days" table) 7334 points but the sum of points from all my machines is only 5928 (according to "Hosts" table).
Why is that?
Have a question about the uptime stated under 'My statistics history' at WCG. I'm running one x4 rig 24/7, but rarely does the uptime equal 4 days. Sometimes it reaches 4:06 or 4:07, other days it will stop at 3:12 or 3:14.
Is this just the way it is or is it something I should look into and try to fix?
Running Ubuntu 10.4, haven't noticed any down time when I've checked the computer itself.
You're good. The daily runtime is based on the WUs that get validated in the day so depending on when you wingmen crunchers return their results you are seeing a little variability which is perfectly normal.
Thanks for clearing that up!
Is it normal that FAAH wu's validate very quickly? I have been running single projects lately and now that I'm doing FAAH I see that they get validated within one day. On boincstats I was seeing 3-4k per day, but the last 2 day it was 5k+. Not complaining at all. But that seems a little high for an i7-860 @ 3.3GHz.
I'm sure this will even out when I'm running FAAH a little longer. Just curious if other have seen this too.
FAAH often has a quorum of one. THe units have proven so reliable that multi-system quorums (quora? quori?) are no longer necessary. They sometimes pull a unit aside for checking but if it doesn't error out, it's more than likely good.
The extra check on a second machine is done on average 20% of the time for ZR (zero redundancy) type WUs.
WCG home forum faq: Lots of answers to lots of questions
http://www.worldcommunitygrid.org/fo...d_thread,21501
noob question but how do I increase additional work buffer to 2 days?
That depends on if you're doing it for just one machine or a whole fleet.
For just one you can either open the BOINC manager and click through: Advanced > Preferences > Network Usage (tab) and then enter the Additional Work Buffer length in days
OR
Go to the WCG website and click through My Grid > Device Manager (it's in the left hand column) > Pick the device profile you want to play with > scroll down to Workunit Cache Settings and put what you want in there.
If you have a bunch of machines and you don't change the individual settings on any of them, then you want to use the second method.
Thanks D A, first one worked fine but the second one I could'nt find the Workunit Cache settings. Have no device profiles other than default so maybee I need to make one for each machine?
Nope, just change the default profile to "custom settings" (the last option in the list of radio buttons) and you can play with most of the settings.
You only get four different profiles (Default, Home, Work and School), but you can have as many machines as you want on each of them.
Got it, thanks again D A :)
Anyone want to help me out getting my sig to work? Dimes isn't covered by other alternative sites, so i'm trying free-dc and ironically enough DIMES are the only stats that show up. Never been able to get this to work, somehow my other projects aren't syncing with it :shrug:
http://stats.free-dc.org/tag.php?name=kingcarcas
The one i have now is fine except i can't get DIMES onto it....
My old folding stats don't have a cross project identifier and the team name is spelled differently so i fear i won't be able to attach those to me :shrug:
Same with DIMES, can you fix this?
hi everyone it sounds stupid but if i run wcg for 24hrs it seems like im getting 3days added to my account i dont know why is it becuse of my cpu been overclocked
ps to get the max points for 1 day what time shuld i start running wcg
If your talking about runtime stats on the WCG page, thats CPU runtime. So a single threaded CPU running for 24h = 1 days worth of CPU run time, an i7 with 8 threads = 8 days of runtime if WCG is run is on 24 hours a day.Quote:
Originally Posted by rubber5
To get the Max points, Run WCG 24 hours per day and use the 64bit client (on a 64 bit machine.) If you have to suspend boinc for work or reasons then just run WCG when you can for the maximum amount of work done :)
i am running 64bit but what i meen about the 24hr is when i run wcg for 24hrs i start it at 12 midnight looking on wcg website it splits the results ie 15000 points then 3000points and so on if you know what i meenQuote:
Originally Posted by trn
i have a pics to show you
(i really dont understand how it works to be honnest)
anyway im i doing good on this as i dont run it 24hrs as you will se by results its to expencive for 1 and 2 its my gaming rig so i do like to play games on it from timt to time
I wanted to say that the simple answer is.... But there is not a simple answer for a number of reasons.
When you run WCG you get credit for each core that you have running and those cores in turn get through a different amount of work depending the speed and type of processor so this results in me, for example, getting 12 years in 2 years.
When you run wcg and complete the work, for some projects at least, you are one of a small number of people doing the same work. This is done to ensure that the results produce good science with no errors. For you to receive credit for the work you do the other people doing that same work must also complete the wu's then they can be compared and proven good. You may, therefore, receive credit some days after you handed in the work. Conclusion: any time of day is good and even if you run 24/7 you will see a fluctuation in the credit awarded each day.
aahhh i know what you meen now thanks for explaning that :up:Quote:
Originally Posted by OldChap
i know im been a pain but i have been crunching for a while and i have got 3 Inconclusive results and im worried becuse i have uploaded 675 results in all and all of them have been valid.
shuld i get worried about this or is it something that happens all the time
thanks everyone
The answer to your question is in the last part of the first post of this thread. I will take a little longer to get validated as the WU was send to an other client.
Ok a quick rundown on inconclusive....This is what happens when results are returned by different machines and they (the results) are different to each other. when there are more than two machines doing the work it will remain inconclusive until all have reported in and the erroneous result can be isolated. where there are only two machines doing the work it will be sent to further machines for comparison when complete.
The short answer is don't worry about these...you only need to worry if your rig produces the wrong result in which case you will find the result listed as an error
thanks for that but if it does return it as an error what will i need to do?Quote:
Originally Posted by OldChap
is it my pc or something else that might have gone rong do you think
thanks for your reponce
The causes of errors are many fold...wait until it happens then depending on the WU or OS etc etc etc we can take a look and proffer advice
ok i will keep you updated and as soon as the 3 have been validated or comeback as error i will let you know as soon as i can :)Quote:
Originally Posted by OldChap
ps i have done a 24hr before and i had no probs then cant see why i will have them now (hopefully)
Ok, if you have one unit that ends up being invalid, it's not the end of the world. Stuff happens. If you get them CONSISTENTLY then you have a problem. Errors are different, they die before the units go back for validation. Those are an issue but even then, one out of the blue doesn't mean much but several in a row means a lot.Quote:
Originally Posted by rubber5
i have just checed this morning and 2out of 3 has been valid it looks like the fault is NOT on my side i thught i was funny me getting them since i have been doint this for a while and i have had no problems so far
thanks to everyone helping me
all of the tasks has now been validated so more worrying :)Quote:
Originally Posted by rubber5
thanks again everyone happy crunching
Is there a known way to use two differentc work schedulers?
Our main work machine is crunching from 5pm till 8am, which is extremely important for fluid svn/web server work from monday till friday.
But Sat and Sun are free days, so it can crunch 24/7. Is there a way for achieving this under linux other than cron requesting a differentc work schedule for weekends and reverting to normal when monday starts?
You can specify crunching and communication times on a per day basis both with the graphical manager application and with the profile settings on the WCG site (via the My Grid pages)
Ok, I just had a play with it and it's not as simple as I'd hoped. Instead I'd suggest running a cron job to start and stop the client at the appropriate times on the relevant days.
Here's a line from my crontab that I use to make sure I don't have finished work waiting to be reported past the daily stats run:
50 9 * * * /home/michael/BOINC/boinccmd --project www.worldcommunitygrid.org update
boinccmd is the cli control utility. It can do pretty much anything the graphical manager can do, though it is a little more effort.
You may or may not need the full path to boinccmd, I use it because I'm like that. The commands you want are "start" and "stop", though you could use "suspend" and "resume", instead of the "update" I've used there and I'm pretty sure you can handle the time entries.
I know the cli drill, we say pretty much 'HELL NO' to any gui on our work servers:)
So basicly, to make sure all goes ok, the standard wcg process will be ended in Friday at 17:01, and after ending it, a new instance of wcg will be spawned - a START one - it'll omit curent work preferences, and will work without any disturbing until Monday, 7.59(here the STOP kicks in), where it's work will end and the standard MON-FRI scheduler(suspen-resume) kicks in. Sounds good.
I'll post the crontab/rc settings when they are confirmed working(and deploy it to our other servers when I finally receive a blessing from the company owners;)
It's always a good thing if you can get that blessing in writing. It's not paranoia if someone really is out to get you. ;)
Rather than run two different clients, I'd be inclined to just start and stop it with cron. That way you'll be less likely to run into issues with work sitting all week waiting to run (all the work we do has deadlines that have to be kept).
I'm not 100% certain but I think this should be all you need:
59 07 * * 2-6 boinccmd --project www.worldcommunitygrid.org suspend
01 17 * * 2-6 boinccmd --project www.worldcommunitygrid.org resume
edit: start and stop might not be the best options unless you prefer to use the rc scripts over the BOINC command tool.
edit2: http://boinc.berkeley.edu/wiki/Boinccmd_tool
So I've started a WCG account a few days ago and joined the XS team.
I have a question.
Now I have joined the Milkyway project too.
Can I make points on Milkyway with my GPU and transfer those points to my WCG account?
And if I can, how do I do this?
Thanks
First of all welcome and thanks for helping!Quote:
Originally Posted by vitchilo
AS to transferring points from MW to WCG the answer is no.
Even though they both run under the BOINC manager they are two separate projects.
hi everyone i have just tested my pc (since i know a little more about wcg)
and in getting about 30.000 points per day with my i7@3.8 ghz
is this about right
i just done 27 hours and i got 33.062 pionts
i would run my pc 24/7 but its not me paying the bill at the end of the day
so i do what i can and so far i have 618.000 points that i put into our team
i know its not much but its better than nothing lol :)
crunch crunch crunch
Hello fellow crunchers. I'm currently crunching on my gaming laptop whenever I'm not gaming and I have an old eMachines desktop that doesn't work anymore however I think I could probably fix it. My question is can I have WCG running on my laptop and my desktop under the same profile at the same time or do I have to make another profile to crunch on 2 computers at once?