Welcome to MilkyWay@home

Posts by MontagsMeeting

1) Message boards : Number crunching : BOINC wins! (Message 26820)
Posted 1 Jul 2009 by MontagsMeeting
Post:
i give up

They have really done everything that BOINC is a pain for me!

  • Not installable on DomainController -> workaround: old version.
  • Not usable on DomainClients -> workaround: unsecure Host so BOINCsecurity works.
  • Cache not per project -> workaround: a lot of timeconsuming babysitting
  • not possible to dedicate project to core/cpu/gpu -> workaround: a lot of timeconsuming babysitting and messed up systems
  • scheduling is a bad joke -> babysitting, babysitting and babysitting


I managed to have a RAC of 130,000 when 180,000 were possible without BOINCs control and 80,000 or less were realistic without my control.

I pay about 200€/Month only for the BOINC part (CPU/GPU at 100%) not including the base costs to run all systems 24/7.
What i get back is the statement that there will be no hope for a better situation as they keep their goal to control everything and let the user as less as possible control over their own systems.

But the reason is not RAC, time or money - it's the feeling to get kicked for my efforts!

I've set all my hosts to not receive new work and in 1 to 20days when the 10day caches are empty, i'll uninstall all BOINC.

Sadly i take 3 users with me as i babysitted their systems what they don't want to do for their own and i suspect that another 2 will follow in a few days when they run into first trouble with BOINc on their own.

Perhaps i'll come back when BOINC gives control over my systems back to me - where it belongs!

Goodbye
Holger

2) Message boards : Number crunching : Configuring BOINC (Message 25810)
Posted 17 Jun 2009 by MontagsMeeting
Post:
If i take control away from BOINC and babysit my PCs very intensely i can do about 180k credit per day. My actual RAC is about 130k, that means BOINC loses 30% and i still have to babysit. If i let BOINC manage things i hardly believe i could keep half of my RAC.
...they know better how BOINC should operate...
Yes, they know it should run half as fast and they do a great job >:D
3) Message boards : Number crunching : Configuring BOINC (Message 25567)
Posted 15 Jun 2009 by MontagsMeeting
Post:
I believe there will be no real solution until BOINC allows us control our own hardware.
For me it looks like the devs are trying to get control over a million different systems running a million different combinations of projects in - guess what - a million different environments with another million goals of their users. I'm sure even to try this, is very stupid!
BOINC should do the basics and let the experienced and interested users enough room to configure BOINC to their needs.
I believe the complete ressource share concept doesn't fit todays hardware. It was OK for single core systems and worked so so on multi-core systems with equal cores/cpus but with the first P4 HT it was obsolete as HT could have a great performance boost if it would be able to assign projects with different cpu-usage to hardware-core and emulated-core. This problem growth the more different the hardware becomes. Actually we have Cores, HT-Cores, cuda, ati and we will become openCL, Larrabee (that perhaps can run cpu-apps and gpu-apps) someone mentioned somewhere soundcards and in worst case all in one PC - as i said, stupid idea to try to control such a system with a standard app - this will be hard enough with very detailed knowledge about the system.

In my eyes BOINCs goals are against their users, mainly the enthusiast users and if i look at the people i personally know and are running BOINC i only know one who wasn't influenced by me, there are five who run BOINC because i do the work configuring/babysitting their systems and another three i told to run it.

If BOINC is continuing to work against me, the point will be reached where i have to say good bye to BOINC taking at least five users with me but rather nine or ten. I would say people like me engage others to join BOINC and making us unhappy isn't a good idea.
4) Message boards : Number crunching : Configuring BOINC (Message 25311)
Posted 13 Jun 2009 by MontagsMeeting
Post:
What is the performance impact of a vm?
i think i have to run the cpu-part on the vm or is the gpu available in a vm? that would be great
5) Message boards : Number crunching : Configuring BOINC (Message 25296)
Posted 13 Jun 2009 by MontagsMeeting
Post:
Today i tried different settings with a ressource share down to 1%, avg_ncpu up to 1 and up to 10 cpus but only a avg_ncpu setting of 0.01 gives me full cpu usage and as Seti has higher priority than MW for me i think i and MW have to live with a hanging MW from time to time.

I tried different BOINC versions too 6.4.5, 6.6.20 and the latest and only 6.6.20 gives usable behavior.

Is there a way to run two BOINC installations? - i think this would solve any BOINC problems.

I think best would be if we would assign projects to cores instead of hosts
6) Message boards : Number crunching : Configuring BOINC (Message 25252)
Posted 13 Jun 2009 by MontagsMeeting
Post:
I try to figure this out for a long time now and with the availability of WUs the situation becomes worse.
I have a core i7 that should run Seti on all cores and a 4850 that should run MW, so far so good my core i7 is running 8 Seti-tasks to achieve this i have to set MW to <avg_ncpus>0.01</avg_ncpus> in app_info.xml. With this avg_ncpus setting every downloaded MW WU will start immediately. On the gpu only 2 of all running MW WUs are really calculated the other >50 tasks stay in memory - with about 15MB per task this makes 700-800MB.
Problem is, that things aren't very stable this way and from time to time MW stops 1 or 2 tasks so i have a lot of active tasks in memory waiting for nothing to happen.

I've tried every setting i'm aware of but nothing heales this behavior
ressource share does absolutely nothing for me
avg_ncpus is the real share-setting anything higher than 0.01 will drop one cpu core after the other
cache-settings too do nothing
7) Message boards : Number crunching : How to Get Work (Message 23785)
Posted 31 May 2009 by MontagsMeeting
Post:
cruncher or muncher :)

8) Message boards : Number crunching : Milestones II (Message 23544)
Posted 28 May 2009 by MontagsMeeting
Post:
BOINC 10,000,000
Milkyway 6,000,000 - RAC 100,000
Seti 4,000,000

all today :D
9) Message boards : Number crunching : Why is it so hard to get work? (Message 23289)
Posted 25 May 2009 by MontagsMeeting
Post:
other than the ATI users, CPU users have many possibilities to crunch on other projects - there is absolutely no way for a CPU to get out of work in the BOINC world.
10) Message boards : Number crunching : Future ATI Card Support? (Message 23136)
Posted 23 May 2009 by MontagsMeeting
Post:
I said that the problem with less credit could happen in the short term and by accident. Despite best intentions and promises, sometimes accidents do happen.
And i have to ask again where is your problem then? And where does it fit all your arguments if all is by accident.

As Paul Buck has said, most likely someone will put effort into attempting to circumvent any block put in place to download work from here...
that would only make sense when credit will be different. For equal credit the non-gpu MW will have much more overhead because of the small WUs so less work can be done and less credit earned. And if less credit happens by accident it will be fixed and no one would like to get WUs from here because it would be stupid for different if not all reasons.

I agree to your first sentence, but that has nothing to do with the bad guys, worst case scenarios you are arguing elsewhere in this thread/board/forum. Things go wrong and if they go wrong they have to be fixed - nothing to worry about.

And to keep it with Murphy again.
Fools are genius, regardless of how much time and energy you spend to make something foolproof, there will be one who invents a way to break it anyway.
Will mean for MW that it is unforseeable what will happen especially how people will react and therefore it doesn't make sense to speculate about it and even less sense to do things based on those speculations.
11) Message boards : Number crunching : Future ATI Card Support? (Message 23092)
Posted 23 May 2009 by MontagsMeeting
Post:
In other words, you're stating that you're going to blatantly go against the wishes of the project?

Of course not, i simply wanted to know if you believe me if i tell complete nonsense, as you ignore every halfway reasonably argument.

I will go with MWgpu because of fewer babysitting, what i stated earlier, even if it would give a little less credit, what will not happen, as travis stated. Looks like you ignore him too if it doesn't feed your paranoia - he said always credit will be the same and you were always arguing less credit this and less credit that ...
You tell things will be lightened up and if taken you by word lightened up things will make the situation worse - that is not conceptual thinking
12) Message boards : Number crunching : Future ATI Card Support? (Message 23036)
Posted 22 May 2009 by MontagsMeeting
Post:
Travis has stated that they intend on not having the GPU apps get credit here once they are up and running over there...

I know how my gpus look like a slightly overclocked dual xeon W5580 >:D
13) Message boards : Number crunching : Future ATI Card Support? (Message 22953)
Posted 21 May 2009 by MontagsMeeting
Post:
If the situation is lightened where is your problem then?
14) Message boards : Number crunching : Future ATI Card Support? (Message 22945)
Posted 21 May 2009 by MontagsMeeting
Post:
Mostly true, yes, but you live in a part of the world that is more socialist than capitalist. Many of the people I'm thinking of will go with the fundamental principles of a Capitalist economy, which is they'll go to wherever they can maximize their profit.
I have two cards that can do more than 140,000credits per day and i get about 90,000 actually
if these cards could get 120,000 credits with MWgpu what would you do? Really stay with MW nongpu for capitalistic reasons???

Another point is that i have to babysit MW a lot and i would love to see WUs that run a few hours for the price of little more real credit against a lot more possible but not available credit - a good capitalistic deal i think - maximum profit for minimum effort
15) Message boards : Number crunching : Future ATI Card Support? (Message 22934)
Posted 21 May 2009 by MontagsMeeting
Post:
Apparently you didn't understand me...or human nature for that matter.

If a GPU task for an ATI card yields less credit / unit time with the GPU project, there will be a sizeable contingent of people who are simply only in it for the points that will stick around here and churn through tasks just like they are today, thus very little will have been gained from the months of work and months of complaining here...

There are two ways for the project to prevent this from happening:

1) Disallow GPUs from processing CPU tasks.
2) Reduce the credit for CPU tasks so that it is not as enticing to GPU users.

If you advocate the second option, you're going to piss off quite a few people.

Alternatively, if people who are in it for the points stick around here because they get more than what they can get there, then CPU users will be shafted again due to the same situation happening with no work availability, so the project may just as well have done what several GPU users advocated, get rid of all of us that use a CPU and go to a GPU-only project. Of course, doing that would've also meant that CUDA would've been first, so there would've been gnashing of fingers that way too, with all the "ATI should've been first"...

I swear...it really isn't this difficult to do conceptual thinking... Maybe it's because I'm a system admin that I tend to look for more than just the one "ideal" outcome. It's called "contingency planning".
I understand your concerns very well, but totally disagree in almost any point.
credit is given by work not time
ATI people won't stay with a project that has no work available when there is a project that supports ATI too even if credit is a little less but available.
If credit will be less for MWgpu this is a calculation error and must be corrected or it is due to a better optimized application then the optimizations should be used with nonGPU MW too.
I think Murphy himself would pale with fear, if he would think in your way about things that can go wrong :D
16) Message boards : Number crunching : Future ATI Card Support? (Message 22910)
Posted 21 May 2009 by MontagsMeeting
Post:
Ice wrote:
That last sentance is correct at least ;) However, BOINC is meant to do a number of things, and one of them is to attract people to donate their computers free of charge. Saying that you need to do no more than contribute your spare cycles is just one way of selling a punter the idea of BOINCing. After that, if you want to run quad cores, and a number of them, exclusively for BOINC, then that's up to you. But certainly no basic mistake.
You're right, it's not a mistake doing it, but as you said it's up to the one who's doing and not up to BOINC or the project.

No-one should be concerned about having no use for a PC (or many) if a projects goes under, as many have. We just switch to another project (or some). No mistake there either for anyone to choose any project and worry about it going bust.
That's the reason for my answer, people here sound like they're very concerned of the work the project has available, so in both points the mistake or better the misunderstandings are the expectations people have and what they think about BOINC/projects expectations.

Sorry i lost the topic a little. I meant, as i said in the first sentence, misunderstandings. English isn't my primary language, so it's sometimes hard to say what i really want to.

So what about ATI cards? Is it a mistake to buy one now? I can't say, I'm unlikely to buy any more, my buying contribution there is done. Was it a mistake to buy ATI cards? I can only speak for myself; I have millions of credits generated in a very short time, and I am turning off electricity hungry CPU crunchers all the time as I concentrate loads'a'crunch into fewer boxes (actually my initial idea was to make do with one box, but then maybe I want more than 'just enough' :)
I know of your other postings that you take the things here as they come and live with it, that's very OK for me, what is not OK for me is that people make decisions for whatever reason and if it doesn't work like they want they search someone/something guilty like BOINC, the projects, ATI-cards, GPUs alltogehter and so on. They only leave out the only one who is really responsible - themselves.

..... and the yellow brick road just seems to go on and on and on.....
sorry, i don't understand this one.

For those who went for ATI at the start I'd say it was no mistake, and for me an interesting journey to get caught up with them in MW crunching. For those thinking about it, you can see what's gone before and what might go from now. As for me, I will have ATI and zillions of credits whatever happens.
for me buying ATI was a big mistake, but not because of BOINC, i'm very happy with having them crunching. But i didn't buy them because of BOINC i didn't know i could use them for BOINC at first. I bought them because of my business where i need the possibility to watch any sort of content. The mistake with ATI is because of their crappy drivers, i have big problems running two 24" Monitors (if that is really the problem, but i've eliminated everything else, i think)


Brian Silvers wrote:
Actually, at least in theory, GPU apps should no longer be able to process CPU tasks once all this is said and done, meaning a CUDA app is available at the GPU project as well as either Brook/CAL or OpenCL. The worst possible scenario is to allow a credit level that is higher here with a GPU processing a CPU task than a GPU processing a GPU task at the GPU project. This would mean that there would be no incentive for people with GPUs to move to the other project and so we'd end up with the heavy consumption of tasks continuing here and little to no utilization of the GPU project.
Why should they drop ATI support? We will see what happens, and then decisions could be made. To make decisions based on some paranoia of worst case scenarios was never a good way. I believe thing will sort out themselves. I don't know if MW can crunch all work it can send only by cpus, if not it would be stupid to drop a second source. I don't know how MWgpu goes, perhaps it's saturated with cuda-cards. Perhaps the best way is to split the available ATI-cards between the projects. And perhaps MWgpu performs that great and uses all available resources that spending resources on a second project isn't possible anymore and old MW will be dropped. I don't even know if it's easily possible to run ATI at the MWgpu as the ATI-app is recognized as cpu-app, there are no ATI-gpu in the BOINC world. Questions over questions that only could speculated about.
17) Message boards : Number crunching : Future ATI Card Support? (Message 22895)
Posted 21 May 2009 by MontagsMeeting
Post:
I understand the frustration of people who invested money into ATI-cards because of BOINC(Milkyway) but in my eyes there are two fundamental misunderstandings

1. BOINC is meant to use spare cpu/gpu cycles to do something useful while not doing their intended work. So any investment dedicated to BOINC is the first basic mistake

2. Neither BOINC nor any project ever promised that any specific system will be supported. Even a complete project could be ended without any entitlement from the volunteers. So to bet on a specific system is the second basic mistake

For Milkyway i don't believe they will end support for ATI not for the non-gpu project and of course not for the gpu project. Cluster Physik already said that he'll work on a ATI app for the gpu project perhaps until the weekend.

I see absolutely no reason for a project to release an app for hardware not supported by BOINC such work will use a lot more project resources than standard apps and is very likely to not work anymore when the hardware will be supported by BOINC, so work has to be done twice. If i take into account how long it took to make the gpu-project, i can't see any free resources for an official ATI-app.

To focus on something not officially supported always includes the risk that it will not work anymore in the future. If you want to go with this risk - do it, if it is too much risk - let it be. But do not blame anyone else for decisions you made for your own.

will all the folks who have purchased ATI cards be left out in the cold?
They will be left exactly there where they were left all the time.
18) Message boards : Number crunching : Catalyst 9.4 question (Message 18836)
Posted 15 Apr 2009 by MontagsMeeting
Post:
I've two 4850/512 with 9.4 under Vista 64 on two i7 920 systems running without any problems as with 9.3, 9.2 and so on
edit: they're running 0.19e with 2 tasks on each gpu, the cpus run seti

the only thing i can't do is to use any remote desktop software, because both systems crash or not after using ms rdp or vnc or teamviewer
19) Message boards : Number crunching : Run the project only a few hours a day (Message 17967)
Posted 8 Apr 2009 by MontagsMeeting
Post:
The correct way to state that is that "no one complains, until the forums are back up and running again"...
I know that this isn't seti. I think it's better to have a defined downtime than the feeling to not get any work because someone else got(stole) it.

What it seems like would happen here is you might get a couple hours of systems getting tasks, then back to where we are now.
That's why it has to be done every day

Of course, since new work is based upon old work, I don't know if it would even be that productive, since there'd come a point where releasing tasks and getting some back would be required...
you might be right at this point, i don't know, but as the actual deadline is 3 days it should easily work 1 day
20) Message boards : Number crunching : Run the project only a few hours a day (Message 17964)
Posted 8 Apr 2009 by MontagsMeeting
Post:
I dont understand where you have your conclusion from, that the server crashed due to the output? I can't see any indication for this. The server worked perfectly for more than two weeks at this output rate.

And i can't see even less what this has to do with my idea, it is completely free of any relation to the output rate. The servers output will be absolutely constant with my idea - to be exact: this is the idea. the only thing that changes is the distribution. Taken your numbers into account there will be only 7 hours needed to do the complete output of one day of the server actually.

And the idea is that it would decrease server accesses as every access is successful while distribution is on and access could be completeley denied while distribution is off and that could even lead to some more free server resources or/and bandwith.


Next 20

©2024 Astroinformatics Group