BOINC

CaptainUnlikely, is this her laptop? http://asteroidsathome.net/boinc/show_host_detail.php?hostid=18890

or is it this one: http://asteroidsathome.net/boinc/show_host_detail.php?hostid=51849

I'm seeing the host in the second link getting this:
<core_client_version>7.0.64</core_client_version>
<![CDATA[
<message>
too many exit(0)s
</message>
]]>

I'm guessing you have used BOINC's Throttling mechanism to reduce the amount of CPU time. If this is the case, I would recommend moving it to 100% and then downloading tthrottle and set it to throttle the CPU when it reaches a pre-determined temp. I have had better luck with this and many others have as well. It is also recommended to just change the number of cores/cpus available to BOINC if you don't want to use 3rd party software. Play around with this option and see if your results change at all.

Another thing is if you have BOINC set to pause while active but have the number of minutes set too short (ex. 3 minutes) you can run into this as well. For example: mouse click, pause, scroll, pause, click again, pause all while reading a web page and spending 3 minutes paused each time can cause BOINC to start and stop a lot in a very short amount of time. I would try setting BOINC to run while active and see if she notices. If she does, you could even use the option to pause BOINC while certain apps are open. Otherwise increase the duration to something higher like 10 minutes. If she watches videos on the laptop maybe 30, 60, or 120 minutes. (That is if you don't set BOINC to pause while those apps are open.)

Another possibility could be if you have the setting for pausing if non-BOINC usage is > then whatever % you assigned is too low.
 
Last edited:
It's the second one, 6930P is my laptop.
I have BOINC set to use 75% of the cores, 100% of the CPU, to help keep the laptop responsive under lighter loads - temperatures aren't a problem, it just felt a little sluggish at times like when watching videos, and I know you'd mentioned setting <100% CPU usage could cause issues so I did it this way instead.
It's not set to pause while the PC is in use, it is set to pause if the CPU usage is over 50%, which I thought would be a decent limit to avoid the situation you described.
I sort of understand what the problem is caused by now though so I'll keep an eye on it, it hasn't errored out before or since so if I keep an eye on it, I should have a rough idea of what the usage was when the errors occurred which should help me sort it out.
 
I typically set mine to >85% since anti virus can spike at times. I also set MS Internet Security Essentials to use no more than 10% so that I don't have to worry. Some pages cause temporary spikes too, so I really don't like the setting. However, on machines that have regular use, I like it just to free up the system when I need it to.
 
So far the Galaxy S3 seems much more responsive than the Bionic. I am running 2 work units at a time. I have only added 3 of the projects for now. I have noticed a big difference in which apps you run and how well the client releases resources.
 
Thanks, testing it out on my laptop now.

Edit: OK, well, first unit I got errored out after about 6 minutes, with the following: "-2147483645 (0xffffffff80000003) Unknown error number"
It's grabbed another unit so I'll see what happens.
 
Last edited:
It is a web crawler app, so light use probably wont be a problem. However, some ISP's don't want web crawlers going on their back bone and if you have a ton of these going, it may catch their attention. Just an FYI.
 
Nothing in their Ts&Cs about web crawlers as far as I'm aware, and if it became a problem I would speak to them about it, they've been pretty good for me. I only have the one client running Plagiarism so far, nearly finished another WU so I'll see what happens when it uploads...finished now, it's waiting for validation. Seems to only download one WU at a time, it doesn't cache any, just uploads a finished unit and grabs one new one. WIll see if this unit validates or not.
 
Most NCI projects only send one. That is where VM's come into play. ;)
 
Gotcha, been a while since I've had an NCI set up other than WUProp since OProject and FreeHAL stopped, and since WUProp "units" are like days long, I wouldn't expect more than one.
Could be good to VM up to the gills along with DIMES then, the process for Plagiarism seems to be using 2.5MB of RAM at present so very light on resources indeed.
 
Last edited:
Wow, miss a week... :D Well, first of all, I just set Plagiarism@home to go, but it's not downloading WUs yet...Do I need to set the Priority higher? I thought NCI projects didn't matter so much when it comes to priority... And second, I became the 17th member of the [H]ard Commandos' 50 million point club! (Yeah, I just made that up...)

Maybe I'll go steal Kendrak's [H]orde Pimp Hat... :D
 
Congrats on the milestone!

I was about to say perhaps they're out of work, they were yesterday, but it seems they have plenty again now. I haven't got any work from them either if that's any comfort heh.
 
They were having problems with the app taking up (locking) a CPU core for no good reason on Windows hosts and allegedly some Linux hosts experienced it. The Admin is more skilled with Linux and so Windows is a little harder for him to work with. So, good to know that there is work again already. Normally things like this takes a few days to straighten out. I'm hoping this one sticks around longer than many of the others that did web crawling. (anansi & surveil)
 
Plagiarism@home keeps crashing out on my laptop, just get a "xxx has stopped working" message. It's doing it fairly regularly, which is quite annoying. Seems fine on my gf's laptop though, not sure what's going on there but I know it's early days so I'm not getting irate yet.
Looks like Asteroids is out of work right now, time to switch! Looks like I'll have a low power home system soon too, picked up an AMD E350 board very cheaply so I'm thinking of throwing together a little file server or something from that.
 
Plagiarism is still Alpha/BETA (the admin has said both) and has had a lot of troubles. So, I wouldn't expect much out of it just yet. I haven't had any that have stated that they stopped working though, so that is a new one to me. I have seen the ones that run forever or keep restarting and also some that just error out. This project is the same admin as Radioactive@home.

I would make that AMD into a HTPC for streaming TV/Netflix/Hulu+/Crackle/Prime/etc... And it should run GPU projects... AMD Radeon HD 6310 graphics
 
Last edited:
Indeed, I know it's early days so I'm expecting lots of bugs. I've had one or two that have run forever, and maybe 5 or 6 that just stop working, plus one that errored out.

Good point on the GPU projects, forgot it had relatively recent onboard graphics so that's cool. I might make it a file server actually, I know it would be good for a HTPC but I don't particularly have a need for one.
 
Congrats on the 50M milestone Jathanis!

Asteroids@home has work once again. Just a minor blip. We only have 5 more days until we catch AMD Users. I'm going to try and throw a couple more cores at it later this week. A few more CPUs on the project and we could really start to make up some ground!
 
Yeah, got one today, and got 11 seconds of run time and this error repeatedly:

11/4/2013 4:16:11 PM | World Community Grid | Task BETA_BETA_9999982_0359_2 exited with zero status but no 'finished' file

I think it's gonna be a bit until they work out all the bugs here... :D
 
Yeah...this BETA is probably the most problematic BETA I have seen released at WCG. I got a hand full of units between the two attempts. Most have errors. Luckily, my netbook got one. So may get some good run time off of that one.
 
WCG BETA seems to be raining right now. Get them while you can. These are the hardest badges to come by.
 
Well theSkyNet explained what happened with the teams when they merged the SourceFinder and the POGS sites/stats together. http://forums.theskynet.org/viewtopic.php?f=2&t=8&sid=3068020bd6cdf3992db27a1db9acde38
It would have been nice to have had a chance to discuss it with the admins before doing it, but that is hind sight now. So, if we want a unified team we need to have everyone at least attach to the POGS BOINC project. phoenicis is the only one that I see that hasn't been moved over yet. If someone could drop him a line to make sure he is aware that would be nice. I honestly don't know much about databases, but I would think the problem would be a lot easier than it is presented. I don't mind if stats are still parsed between the two for stats sites, but the teams should still be unified. It was mentioned that someone might want to be on two different teams (one for each client), but I don't see any other projects with sub-projects doing this.
 
Last edited:
Just thought I would give a quick update here on me & my gf's situation, and explain that that's the reason why I haven't been particularly active as neither of us are getting much sleep at all these days, and money is getting pretty tight too as we're basically living off 1 income now and she's been unable to work for the last 15 weeks or so.

We had her results back eventually, the MRI showed no abnormalities, the EEG was basically inconclusive. They want to keep her on the tablets she's taking in case it is epilepsy or close to it, and keep slowly increasing the dose. She's been on them I would say 5 weeks, no signs of improvement, and they make her even more tired, but she's persevering.
Last night she had a really bad episode so I called an ambulance again, they took her down to A&E. On a Friday night as you can imagine it was incredibly busy...we got there around 10pm at a guess, saw a doctor about 11.30 who was really shocked at the results so far and that the neurology department hadn't come up with something yet. They asked a lot of detailed questions and seemed pretty helpful but by this point my gf was incredibly tired and it was getting difficult to answer everything and respond properly. They went away, and about 1am someone else came to say she'd be kept in overnight, which made her quite upset as she just wanted to go home and sleep - understandably. By about 2am they finally moved her to the observation ward, and 2.30 or so I was on my way home - up for 22 hours on a diet of a few hours' a night sleep, I was just about seeing things at this point heh. She called this morning and said they're keeping her in for observation til 8pm tonight so I'm just drinking the largest coffee I can make before I head down there with all her stuff.

So, not going too well overall...it's getting more difficult day by day for her, and in all honesty we just don't know where to turn. She hates being stuck at home on her own all day every day, we're both incredibly tired and drained all the time, and like I said money is running really tight, the government have refused her any help with benefits and her sick pay from work (which just abut covers our food each week, not much more) potentially ends in a week or so, at which point we will be living off one sole income. Leaving money aside, as I said she's stuck at home, can't go out on her own even to catch a bus, can't really even have a shower or cook due to the risk of having a seizure, so really all she can do is sit in bed or sit on the sofa, watch films, maybe play games, read...all well and good for a week or two but not after this much time.

In terms of BOINC and computery things, I have very little time and even less energy & motivation to work on any of my projects, and if her sick pay does end it's likely I'll have to sell off my unfinished SR-2 build and anything else I can, I might also have to sell my car although that's one thing I would prefer to keep in this situation. Hopefully, things don't get that bad, still keeping my fingers crossed they will be able to help her soon because it's really destroying her life.
 
Sorry to hear the troubles Captain. We are certainly keeping the two of you in our thoughts and prayers. Keep us updated when you can.
 
Thanks Gilthanis, it's much appreciated.
They said yesterday afternoon they would keep her for a further 24 hours, I was there from about 11am to 5pm. She's been moved to another ward with more restrictive visiting hours so I'm heading there 2-4pm today. She should be let out between 2pm and 7pm, hoping earlier rather than later, then it's a trip to TGI Fridays to try and cheer her up as much as possible, it's been a very rough weekend for her so far.
 
Gilthanis said:
Hey metallicafan, I see they have the Christmas Challenge listed. http://www.worldcommunitygrid.org/team/challenge/viewTeamChallenge.do?challengeId=6302

Hopefully our FAH brothers meet their goal quicker so that we can get more of a showing from them this year. Last years boost was due to GPU support of HCC. We don't have that this year. Not unless they (WCG) pull something huge out of their butts before then.
Click to expand...

I do not see [H] on that list, when does it start ;)
 
It starts December 1st.

Edit: update on theSkyNet : http://forums.theskynet.org/viewtopic.php?f=2&t=2778 So, if someone could contact phoenicis to ask him to join the POGS team, we can have a smooth transition to a single unified team. He was the only one that I saw who was on the Source Finder side but not on the BOINC POGS side. The link mentions that if users aren't on the BOINC team when it gets merged, it could create problems.
 
Last edited:
Berkeley is now also offering a download option that includes virtualbox. This is handy, but I have read the installation may be a bit confusing. Evidently, the version with VB says that VB is necessary and then the next screen says it is optional. I think most people familiar with BOINC will figure it out. :) http://boinc.berkeley.edu/download_all.php
 
Totally random observation - I have POEM GPU set up to run 4 WU at .75 CPU each (anything less than .75 caused run times to increase), so I can still use one core of the quad my 680 is in for other stuff. So this weekend we took my daughter to Disneyland, and I started F@H for the Overtake Default push (and because that 8.4M score I had really needed to be an even 10M, lol) but forgot to set POEM to NO NEW TASKS. Come home to find 4 straight days of 140K to 200K ppd in POEM, while F@H is still running and getting about 30K ppd at the same time...

So apparently, I can run F@H and POEM GPU at the same time with no issue! :D I wonder anout how much POEM running takes away ppd from F@H, as the POEM ppd is right on par with what I've read for a maxed 680...
 
That is awesome. That 680 has really boosted your production. :)
 
I hit my SETI BETA 1 million point goal today. That makes project #21 :)
 
Last edited:
You must log in or register to reply here.
Back
Top