DC Vault 2

If I can make it work properly I'll be setting up a client for MJ-12 at home tonight, my connection seems fine during off-peak hours just not during peak. Not ideal and believe me, they WILL be sorting this, but I'll at least be able to have it crawl overnight for a good 8 hours or so which should help.

I've set my clients to no new work as I was mainly pushing NFS, I was thinking Climate Prediction might be fun. IIRC it's not multi threaded though and I know it can eat disk space, slightly concerned that some of my systems might easily fill their disk space with 8 or 12 threads running and it then means setting separate preferences to each depending how much space each thread takes. Plus, as you said Jathanis, if it might be removed and the stats are a mess that could be a bit of a waste, I'd rather let things settle and see how it goes before hitting it.

Sounds like Yoyo might be worthwhile then?
 
I'm watching CPDN to see if the stats truly are fixed or not. They seem to be working now, but they aren't the greatest at resolving things quick. The work units seem to run solid now. a few months back it was a mess. The problem though is that they have a work unit shortage right now and are trying to fix that. So, I don't know how well a push there would be yet. If they pull it from the Vault which is less likely now that their stats are "fixed", then it is a waste in a way. However, it still advances us in Formula-BOINC which we currently have no points for.

YoYo is a nice choice, but I would be limited on my support there since I hit most of my goals already. I would suggest if YoYo to do a double push. YoYo on CPU and Asteroids on GPU. However, that is nVidia only.

SOB would be a nice push, but my setups probably wouldn't push very many work units with a months push. Probably just finish one work unit per thread/core and then the second work unit would kill the following months push. I do have a couple cores at work dedicated to it though. I have also not tried running the client in Linux. So, someone who has may want to make a Linux guide that is Grandpa proof.
 
So I guess we have a fun matrix of projects and requirements :)

Boinc
CPU
MOO/RC75?
YoYo:​
GPU
nvidia Asteroids:​
amd Moo:​

Non-Boinc (more complicated setup)
CPU
SOB: NickOfTime, tjmagneto, Gilthanis​
DPAD: Dmaine​
GPU
GIMPS: NickOfTime, brilong​
BANDWIDTH
MJ-12: CaptianUnlikely, DooKey, Jathanis​
 
Last edited:
I'm just as much a Linux noob as gramps.

But I will have to figure it out since I'm running Linux and Windows.

Well, I started it on a linux box but didn't set it up to start on boot....
Rough Guide...
Download MPrime. & Create an Account http://www.seventeenorbust.com/download/
Unzip MPrime
Edit Prime.txt or download from SOB/Downloads while logged in...
Open Terminal and Navigate to Directory
Run Command ./mprime -m
It will communicate with the server and Start processing based on defaults​
Configure overall Memory usage with 13. Options/CPU
Check Worker Status with 3. Test/Status
Adjust Worker Thread setup with 2 Test/Worker threads
Start and stop theads with 4
Exit Client with 5

After Exiting with 5 and restarting you may have to select 4 to continue...
 
YoYo appears to have some app problems on my MP rigs, while it may not show up on 1P rigs it appears to show up on MP rigs. It seems to have a memory leak or something along those lines I have not investigated the issue nor am I going to. I do know that on the rigs I was running yoyo on, when I switched them over to the primgrid challenge they preformed terribly until the first WU completed no matter how many times I rebooted.

Once the first primgrid WU fished they ran fine, but the first WU took 2x longer than it should have. This only happened on the rigs that were running yoyo prior to the primgrid challenge starting, I therefore believe it has something to do with yoyo. The rigs that were running other projects had no problems what so ever. I most likely will not be inclined to run yoyo at this point in time.
 
So I guess we have a fun matrix of projects and requirements :)

Boinc
CPU
MOO/RC75?
YoYo:​
GPU
Asteroids:​

Non-Boinc (more complicated setup)
CPU
SOB: NickOfTime, tjmagneto​
GPU
GIMPS: NickOfTime, brilong​
BANDWIDTH
MJ-12: CaptianUnlikely, DooKey, Jathanis​

Moo! is GPU as well, and since Asteroids is nVidia only, you might want to list it as the AMD alternative. I want to again remind anyone running Moo! to follow these directions: http://hardforum.com/showpost.php?p=1040794811&postcount=1947 And if you run a AMD 7970 on this project, please read this thread first: http://moowrap.net/forum_thread.php?id=200 There is a link in it for the files you will need to get it to run correct. I have used it and can confirm it is necessary.
 
My suggestions for addition:
Constellation
Ibercivis (they gave permission via PM)
PrimaBoinca - only concern is lack of forum response from mods/admins
DistrRTgen
Numberfields (they gave permission via PM) - they also said they have multiple years worth of work left to do.
Distributed Data Mining - users should take note that it requires Java to be installed

Maybe for next month/summer we can start booking more points for the projects suggested for additions :)
 
Rusty posted to the admins that he has given us enough time to get a good list. So, he is planning on doing a "public" discussion on it next.
 
Attempt at DC Vault Summary....

Physical.Science

...................................Rank. Pts/Rank..Next.....Stats
Asteroids@home..........26......11.29....2/10d....http://stats.free-dc.org/stats.php?page=team&proj=ast&team=1620
.............................................................4/90d
Climate.Prediction......265......1.29........lots.....http://stats.free-dc.org/stats.php?page=team&proj=bcpdn&team=10334
Cosmology@Home......37......5.23........51d.....http://stats.free-dc.org/stats.php?page=team&proj=cos&team=1131
Einstein@Home..........123......0.92.........-........http://stats.free-dc.org/stats.php?page=team&proj=eah&team=1254
eOn.2..........................14.......14.58........R......http://stats.free-dc.org/stats.php?page=team&proj=eon2&team=888
Leiden.Classic............28........7.73........12d....http://stats.free-dc.org/stats.php?page=team&proj=lei&team=1027
LHC@Home...............89........2.07..........R......http://stats.free-dc.org/stats.php?page=team&proj=lhct&team=591
MilkyWay@Home.......58........2.62........28d....http://stats.free-dc.org/stats.php?page=team&proj=mil&team=1011
Muon1.DPAD..............9........135.14........-......http://stats.free-dc.org/stats.php?page=team&proj=dpad&team=[H]ard|OCP
Seti.(BOINC)..............214........0.17......7/90...http://stats.free-dc.org/stats.php?page=team&proj=sah&team=30839
theSkyNet.POGS........9.........19.09.........-......http://stats.free-dc.org/stats.php?page=team&proj=pog&team=1684.

Biological.And.Medical.Science

...................................Rank. Pts/Rank..Next.....Stats
Docking@Home..........20.......8.61..........R.......http://stats.free-dc.org/stats.php?page=team&proj=doc&team=330.
Folding@Home............1.........0.12
GPUGRID...................26........7.42........120d....http://stats.free-dc.org/stats.php?page=team&proj=ps3&team=799
Malariacontrol.net........61.......4.31..........92d....http://stats.free-dc.org/stats.php?page=team&proj=mal&team=1845
POEM@Home...........42........6.30..........23d....http://stats.free-dc.org/stats.php?page=team&proj=poe&team=825
[email protected]......19.89...........R......http://stats.free-dc.org/stats.php?page=team&proj=rna&team=773
Rosetta@Home..........62.......0.97.........5/31d..http://stats.free-dc.org/stats.php?page=team&proj=rah&team=5560
SIMAP.........................23........4.09............-......http://stats.free-dc.org/stats.php?page=team&proj=sim&team=4227
The.Lattice.Project.....28........9.22...........R......http://stats.free-dc.org/stats.php?page=team&proj=lat&team=1187

Mathematical

...................................Rank. Pts/Rank..Next.....Stats
ABC@Home..............51.........5.40...........R......http://stats.free-dc.org/stats.php?page=team&proj=abc&team=1133
Collatz.Conjecture....54..........6.71.........25d.....http://stats.free-dc.org/stats.php?page=team&proj=col&team=594
Enigma@Home........34.........7.71.........2/45d...http://stats.free-dc.org/stats.php?page=team&proj=eni&team=715
GIMPS......................14........18.87.........2/31d..http://stats.free-dc.org/stats.php?page=team&proj=gimps&team=342
...............................................................7/90d
NFS@Home............16........15.09.........2/15d...http://stats.free-dc.org/stats.php[/email]?page=team&proj=nfs&team=613
...............................................................3/90d
PSP.-.PRP................5.........333.34.......2/50d...http://stats.free-dc.org/stats.php?page=team&proj=psp&team=[H]ard|OCP
PrimeGrid................52...........3.64........3/10d...http://stats.free-dc.org/stats.php?page=team&proj=pgrid&team=1710
..............................................................12/93d
RC5-72...................74............2.02............-......http://stats.free-dc.org/stats.php?page=team&proj=rc572&team=HardOCP
................................................................MOO.http://stats.free-dc.org/stats.php?page=team&proj=moo&team=1389
Seventeen.Or.Bust..28.........17.40..........68d.....http://stats.free-dc.org/stats.php?page=team&proj=sob&team=646#
...............................................................4/90d

Miscellaneous

...................................Rank. Pts/Rank..Next.....Stats
Dimes.......................9...........5.93...........49d....http://stats.free-dc.org/stats.php?page=team&proj=dimes&team=[H]ard|OCP
Majestic-12...............7...........-26.........37d....http://stats.free-dc.org/stats.php?page=team&proj=maj&team=147
SZTAKI.Desktop.Grid.21.........6.23...........-.......http://stats.free-dc.org/stats.php?page=team&proj=sztaki&team=1453
World.Community.Grid..57.....0.33..........83d....http://stats.free-dc.org/stats.php?page=team&proj=bwcg&team=613
yoyo@home..............23........11.12...........-......http://stats.free-dc.org/stats.php?page=team&proj=yoy&team=407

Addition

..................................Rank.........Next.....Stats
Constellation..............NA
Ibercivis.......................253.......-.......http://stats.free-dc.org/stats.php?page=team&proj=ibe&team=1501
PrimaBoinca...............132...5/45d...http://stats.free-dc.org/stats.php?page=team&proj=pri&team=84
DistrRTgen...................83.......-.......http://stats.free-dc.org/stats.php?page=team&proj=drt&team=149
Numberfields................55.......-.......http://stats.free-dc.org/stats.php?page=team&proj=num&team=19
Distributed.Data.Mining.47...2/41d...http://stats.free-dc.org/stats.php?page=team&proj=ddm&team=948

Estimated Points (based on current activity)
400 / 30d
950 / 90d
 
Last edited:
Ibercivis has a server issue that they are presently working on for uploading work and they can't fix that while the others are on. So it is just a waiting game.
 
We overtook the Knights in the Math category for position 4.

I have also added 3 more ARM devices that were each single core 1GHz for part time crunching. This should also help our Formula BOINC position with QCN. We are currently 5th place for that project there. With OProject adding ALX back, we should start getting a boost there which should help too.

NickOfTime, thanks for the run down. I know that takes time. I will mention though that the link for Majestic-12 is wrong. It points to DIMES stats. I nominate you to be the keeper of the breakdown... :D
 
Last edited:
With OProject adding ALX back, we should start getting a boost there which should help too.

Somehow I missed your post on this, but I added it to both my Android devices...
So if I have this right, it will run 16 second WU at a time, then wait several minutes to upload and get another? Any way you know of to speed this up on NativeBOINC, Gilthanis?

Edit: Just for the hell of it, I pulled up an Android emulator I had floating around for Windows, installed NativeBOINC, and right away it downloaded and started running Enigma and WUProp - still waiting for it to try OProject. In theory, one could run numerous instances of this via VM for the NCI projects, as the total frontend, agent, and various services all take up less than 200MB... Not that I'm in a big hurry to try and tie up that much core power/memory - the BlueStacks/NativeBOINC instance it using one entire thread (12.5%) of my i7! I'll run it for a day or two and report back where half a hyperthreaded i7 core @ 3.8 GHz compares to my Androids...
 
Last edited:
I have messed with VM's in the past for Android. However, I have not yet played with BlueStacks even though I was in a discussion on IRC about it. And yes, Razor_FX_II ran a ton of the ALX last time in VM's and cranked out a ton. However, right now I think there is limited work units as the admin tests the new feeder. The admin bites off more than he can chew and will let things go unattended for long periods. So, get them while you can. Who knows how long they will last. What he hasn't said is if the application does anything yet. Before, it did nothing. So, I don't know if he is just using it to test delivery capable or what...
 
If I can make it work properly I'll be setting up a client for MJ-12 at home tonight, my connection seems fine during off-peak hours just not during peak. Not ideal and believe me, they WILL be sorting this, but I'll at least be able to have it crawl overnight for a good 8 hours or so which should help.

So it seems there's a network issue in the are, expected resolution date is approx 2 weeks. They reduced my bill by 25% for the next 18 months and gave me a free early upgrade from 60mbps to 100mbps, so I'll see how it goes, and once it's back to working properly I'll get MJ-12 going. I currently get around 6-7mbps but it's very variable, so not much point right now. I also ran into issues getting MJ-12 installed on Debian, seems it has an old version of glibc and I tried to update it but either failed miserable or it still didn't work - not sure if I messed up or it just isn't right. Either way it'll need reinstalling, I'll try Ubuntu as MJ-12 worked first time on my Atom running Ubuntu and I'm just not good enough with Linux to spend too much time troubleshooting.

I attached my hosts to OProject, thanks for the heads-up. I'm all over the place with work again this week but I'm hoping to get another hex up and running by the weekend. If the heatsinks I bought actually fitted properly it'd be a good start :rolleyes: see if I can dig out something that'll work as the backplates on the coolers I had that I could easily get at seem to interfere with some components on the underside of the board so the backplate doesn't sit right which is irritating.
 
I don't think that Rysiu made any real changes to the ALX application, so it is probably just being put out to test his new work generator. So, keep an eye on it. I remember some people ran into times where it locked an entire core even though it wasn't using it. I didn't see it much at all, but others complained. Also, if it is the same application, it does absolutely no science. So, run it only if you want to generate points for a personal goal or for positioning in Formula-BOINC.

Also, in regards to Majestic-12, every little bit counts. Might as well run it and get the kinks out now and then open the flood gates when your speeds are better.
 
It seems to be behaving itself so far but I'll keep an eye on it. I was mainly thinking of our position in FB, if it helps then it's worth running, unless it does start to cause issues.
I'll see if I can get that machine reinstalled tomorrow night and have another crack at MJ-12, it'll be a good opportunity to test the profiles feature actually - to run it only overnight or when I'm not at home. Just another one of those crazy weeks where I'm all over the place driving between offices so by the time I get home I just want to sleep, but hopefully I'll be home a little early tomorrow to give that machine some time.
 
It seems to be behaving itself so far but I'll keep an eye on it. I was mainly thinking of our position in FB, if it helps then it's worth running, unless it does start to cause issues.
I'll see if I can get that machine reinstalled tomorrow night and have another crack at MJ-12, it'll be a good opportunity to test the profiles feature actually - to run it only overnight or when I'm not at home. Just another one of those crazy weeks where I'm all over the place driving between offices so by the time I get home I just want to sleep, but hopefully I'll be home a little early tomorrow to give that machine some time.

By the way, how's the ol' lady?
 
By the way, how's the ol' lady?

Not too bad, thanks for asking, but she wouldn't thank you for calling her the ol' lady ;) thankfully she's staying with her family for a change of scenery at the minute so she won't see heh, she's back Sunday,
She's finished her course of therapy, no real improvement in her seizures but it has helped in general. We have another neurologist appointment mid July, I think she'll be discharged as non-epileptic seizures don't really come under neurology and they're pretty sure that's what it is rather than epilepsy. Not really sure what their next step is to be honest, the government are dragging their feet deciding if she needs any kind of assistance which is nice of them, I mean clearly she should just go back to work or something when she can't even safely leave the house by herself. At this point she's been unable to work for 12 months and being stuck inside, alone, is driving her absolutely crazy.
 
It appears that custom clients are still being used at least by a few at POGS. I honestly don't know why this can't be resolved already. If Kevin is away so much, one of the others could at least address the issue more than chime in that Kevin is away. At least most people respect the project and honor their wishes. I for one would not mind to see drastic measures at this point. I'm sure wiping that user's points entirely would get them to chime in. This is one of the reasons why I don't understand the reasoning for giving people the option of hiding their systems. I don't see any positive reason for this. To me if you hide your systems, you are doing it to hide your activities. Hiding implies negative activities. I know this isn't the case most of the time, but just don't get the reasoning why someone playing by the rules would ever need to hide their systems to begin with.
 
I just had a Yoyo unit running a fair way past 100% so I aborted it. Also, I have two OProject ALX units running at once on this rig, seems odd? Neither is using any CPU which is how it should be, but it does seem weird that it's running two simultaneously.

Edit: they are both locking a thread though, which I assume is because BOINC doesn't know they're NCI. Also I have a Yoyo work unit consuming 1.1GB of RAM whilst the others are all using 1 or 2MB, do some Yoyo units use that much RAM? Hmm in fact now they're all around 2MB, that 1.1GB one has disappeared. Not sure what's going on here, investigating.
 
Last edited:
Interesting, currently have 8 ECM units running on this rig and they're all around 2MB, maybe it just spikes at odd times. It was an ECM unit at 1.1GB though, so at least it's not entirely unexpected, thanks for that.
 
Boom goes the dynamite! We move up a position in PSP this evening.
 
I just had a Yoyo unit running a fair way past 100% so I aborted it. Also, I have two OProject ALX units running at once on this rig, seems odd? Neither is using any CPU which is how it should be, but it does seem weird that it's running two simultaneously.

Edit: they are both locking a thread though, which I assume is because BOINC doesn't know they're NCI. Also I have a Yoyo work unit consuming 1.1GB of RAM whilst the others are all using 1 or 2MB, do some Yoyo units use that much RAM? Hmm in fact now they're all around 2MB, that 1.1GB one has disappeared. Not sure what's going on here, investigating.

I am pretty sure that is some of th memory leak problem I encountered with yoyo when you run lots of them at times there will be several of them running 1.1GB of ram and when they are running the computer get sluggish even if you have plenty of excess ram and I did notice the others ram usage going up when they were running. In general I believe yoyo as some software problems and is not the best written app. :(
 
Boom goes the dynamite! We move up a position in PSP this evening.

Damn I guess I better go see if they added me to the team but from what I saw they do not need my help there, it will not take them long to get to #1 :eek:
 
Damn I guess I better go see if they added me to the team but from what I saw they do not need my help there, it will not take them long to get to #1 :eek:

I didn't see you in the Free-DC stats yet. Make sure to let me know how long it takes total to get you added. It is crap to take this long to add someone to a team. :rolleyes:
 
Big props to brilong in WCG. Way to go sir. Thanks for the very impressive contribution to the team and the science.

Speaking of WCG - I have been thinking about running some CEP2 but in the past I have heard some speculation/comments about how hard it is on HDD/SSDs. I have plenty of RAM in my 2500K box but do I need to worry about the disk? Anyone have experience here? Thanks.
 
6 weeks less 2 days... For psp team assignment.

I think at Brilong's last checked rate it was 45d till the next rank... And then a really long while till the next one...

I see that eon is still around... 7 day removal rate huh
 
Last edited:
Speaking of WCG - I have been thinking about running some CEP2 but in the past I have heard some speculation/comments about how hard it is on HDD/SSDs. I have plenty of RAM in my 2500K box but do I need to worry about the disk? Anyone have experience here? Thanks.

I have two Ubuntu Server boxes with old WD/Seagate hard drives dedicated to CEP2 and I'm not having any problems running 4 units simultaneously. I had nothing but problems on Windows with newer hardware. It seems CEP2 is much more stable on Linux than Windows. File uploads are 20MB right now. My trusty WRT54G is handling them fine.
 
metallicafan
I only had issues with CEP2 on shit ass Hitachi branded drives and drives that were starting to go anyways. Other than that, I didn't have much trouble. Obviously laptop hard drives are slower, so you may not want to push those so much. I used to run up to 8 on my i7 920 but it has been a long time since I ran more than just one per system.

NickOfTime
I will be bringing this up with Rusty. I think these kind of things should be squashed since they are within his direct control.
 
Last edited:
6 weeks less 2 days... For psp team assignment.

I think at Brilong's last checked rate it was 45d till the next rank... And then a really long while till the next one...

I see that eon is still around... 7 day removal rate huh

There is one guy "Lars" (ltd) which handles the team assignments. If he's away on work assignments, he cannot add people remotely, so he has to go home and add them. It's not ideal, but I think the pain might be worth it if we can move up the ranks in DC-Vault!
 
Due to more and more of our FAH brethren wanting to switch to similar science as FAH in the BOINC world, I have suggested GPUGrid quite a bit. I recently requested more CPU work in their forums (since it seemed to be hard to come by) and sure enough they added a few thousand. I don't know if it was coincidence or not, but it happened. I think we should put in some more time on this project. I know the GPU's dominate, but with the number of cores we have, we could probably keep their feeder dry of CPU work....lol.
 
For Protein related projects your pretty much limited to GPU Grid(GPU and CPU), Poem(CPU), and Rosetta(CPU) for getting a steady supply of work. Plus Poem is looking to finally get the bugs ironed out of their new GPU application so they will have GPU work available sooner or later.

Then there is CAS and NRG (both CPU) if you don't mind spells of no work more often than not.

Until the end of the year, when they are going to stop putting out WU's to the public and go all in house computing, there is also Simap(CPU).

I don't think there are any other projects out there with work directly related to Proteins beyond that list.
 
<snip>Poem is looking to finally get the bugs ironed out of their new GPU application so they will have GPU work available later or even later.

There, FIFY! :D

Seriously, I'd LOVE to see POEM get their issues worked out, but after the sheer number of false starts and error filled test runs they've had, I have little faith they'll get things fixed any time soon...
 
Due to more and more of our FAH brethren wanting to switch to similar science as FAH in the BOINC world, I have suggested GPUGrid quite a bit. I recently requested more CPU work in their forums (since it seemed to be hard to come by) and sure enough they added a few thousand. I don't know if it was coincidence or not, but it happened. I think we should put in some more time on this project. I know the GPU's dominate, but with the number of cores we have, we could probably keep their feeder dry of CPU work....lol.

Nate confirmed in their forums on my post that he added more work. Last I checked, they had ~4k work units for CPU work ready to go.


Edit: As of now, the work units are up to over 27k to send out.
 
Last edited:
The admins are trying to wrap up the discussion on additions and removals at the moment. Just a few more on-the-fence projects to agree to. Then we will be posting them in the public forums for anyone else to stake their claims. Hopefully, this happens soon.
 
There, FIFY! :D

Seriously, I'd LOVE to see POEM get their issues worked out, but after the sheer number of false starts and error filled test runs they've had, I have little faith they'll get things fixed any time soon...

I've been waiting on this as well as RNA World to get their stuff working. They require Virtualbox for their latest WUs and I've got it installed on a bunch of my machines, but no work units have been sent down yet.
 
I've been waiting on this as well as RNA World to get their stuff working. They require Virtualbox for their latest WUs and I've got it installed on a bunch of my machines, but no work units have been sent down yet.

RNA is like hitting the lottery. However, I have one in progress right now. And it will be a while: Intel(R) Xeon(R) CPU E5-2650L 0 @ 1.80GHz [Family 6 Model 45 Stepping 6]
 
As expected, Collatz is down. The admin is on vacation, so anyone trying to push this project should just run something else for a bit.
 
Nate confirmed in their forums on my post that he added more work. Last I checked, they had ~4k work units for CPU work ready to go.


Edit: As of now, the work units are up to over 27k to send out.

I wonder if they could keep up with brilong, Grandpa, and DooKey at the same time hammering for CPU work. hmmmm.... :D
 
Back
Top