Posts by KAMCOBILL
log in
1) Message boards : Number crunching : Credit Calculations. (Message 4481)
Posted 25 Jul 2008 by KAMCOBILL
I agree with P4 getting more credit for the WUs they run. That's what's fair about. A CPU second is a CPU second. The P4 runs 1 WU and the Quad will run 4 WUs. Same Credits for 8 Hours of work.


That's like saying 2 Sprinters should get the same Gold Medal even though 1 of the Sprinters who was older & slower only ran 1 Lap while the other younger & faster Sprinter ran 4 Laps of a Race.

In your View since they both spent the same amount of time on the Race Track then they both should receive the same Reward even though 1 ran 1 Lap & the other 4 Laps ... ???


It depends what the goal of the race is. I thought the goal was to do as much as you can do to the best of your ability. Not to be the first one finished or highest scorer. It's a team effort of all the BOINC Crunchers to make all the projects a success.

As I said I'm not credit oreiented but it seems all the Slow computer aren't wanted as much as the Fast ones.

It seems like the projects give high or low credits or just right credits don't have anything to do with type of computer or how fast they complete the task.

Hehe, we're in it together.
2) Message boards : Number crunching : Credit Calculations. (Message 4479)
Posted 25 Jul 2008 by KAMCOBILL
I agree with P4 getting more credit for the WUs they run. That's what's fair about. A CPU second is a CPU second. The P4 runs 1 WU and the Quad will run 4 WUs. Same Credits for 8 Hours of work.

Just like in your example that get the same amount of credits. Parity. If someone runs their P4 24 hours a day and someone runs the quad 24 hours a day, It's should be the same. I have PII up to Quads running 24/7/365. In most cases the electric cost, time spent on each for maintenace and CPU seconds are the same.

Now, I have to say this as a non-bias cruncher. I'm not into it for the credits. It's is a nice tool to see how I compare to others though. I'm attached to 83 projects and they all get the same amout of resources, except for the single platforms that I don't run and the projects that won't give WU to host (e.g. lower memory, older OS version and etc. I've been attached to QCN Alpha Test Project since it went live which it's WUs aren't CPU intensive and don't give any credits.

Which brings about the only flaw in the system, maybe. How to credit it and DepSpid?

I would say that someone that is in it for the credits are going to run the higher credit giver projects and/or give more resources to them. Someone in it for the good of the world or hobby would run them all.

As for SAH, nothing against it, loosing a lot of CPU time from me by "WU cancelled by Server" Even on my Quads. That tells me that if your not fast enough or if your running other projects, we don't want you. Of course, that not the only one I get WUs cancelled on, there are a couple of others I get it.

Last but not least, It don't matter what the credit system is, if Users/Teams have the resources and use them they'll still be in the same position in the project that they are in. So I cannot see the Fuss. Think about it.

3) Message boards : Number crunching : Credit Calculations. (Message 4475)
Posted 25 Jul 2008 by KAMCOBILL
The only fair credit system that will work to make all project credits, reward for work & computers equal/fair is 1 CPU Second = 1 Credit (or any fixed credit per CPU Second). This will make all projects and computers equal if that the goal being seeked.
4) Message boards : Number crunching : Credit Calculations. (Message 4387)
Posted 22 Jul 2008 by KAMCOBILL


I'm not sure what happen in Nov 2006 besides one of my hosts crashed and then the problem bagan. All my old projects had there CPIDs stolen away in this transition. :D What I can't figure is how can a host have different CPIDs (4 in most of them) and control/change the Project CPIDs when they connect. I thought the Projects Servers controled the hosts. The one good thing is I got down from 10 CPIDs to 4. I even detached from BAM to see if CPID Problem would clear up but it didn't

When BAM retires a project they detach all the hosts from the project and it keeps detaching them for an eternity. An example is ABC@Home Beta. It only has work as needed (Nov 2007 was last WU I finished) & it wasn't reachable for a while so I guess that is why they retired it. I attach to it ever once in a while to see if it up and as soon as my host syncs BAM it detached. If It does give out any work, I'll have to Detach from BAM to run it.



I had the same Problem for about 2 Years myself, my Stats would jump up & down at the Stat's Sites



Hmmm...

Now that is interesting. I seem to remember loosing a couple of host CPID's on BOINCStats around then, but everything resync'ed on the same Account CPID.

The only side effect was a couple of hosts lost their graphical data for everything from before the event.

Let me check my logs and get back to you on this.

Alinator



I tried everything to solve the problem. Including begging the Dev (which is a low priorty issue). I would like to see if they could straighten out the problem, but I think I going to detach 15 hosts from all the projects, delete/clear out all BOINC. Then go into one node and change all the CPID to my main one (or suspose to be my main one) then run it until I get all the projects to Sync with my host. Then put the others back online again.


I had the same problem for about 2 Years myself, my Stats would jump up & down at the Stat's Sites because the Projects I ran would fail to keep the same CPID across them. I tried everything from editing the .xml files to trying some suggestions JMV gave me which didn't work either.

What I finally discovered was that 1 Project was the key to keeping the same CPID at all the other Projects, that Project for me was Pirates@Home. As soon as I attached Pirates@Home to all my Computers the CPID straightened itself out across the Projects and to this day I don't need to have Pirates Attached anymore but do keep it attached to 1 Box just for old times sake ... :)

So maybe thats what you need to do, find the Key Project, it's more than likely 1 of the first Projects you ever ran as Pirates was for me, it's something to think about and look for anyway ... Good Luck


That's what I can't figure out, they are attached on all hosts. The only one that is guestionable is Predictor@Home. It was down for a long time, but has been back for a while.

Then there's projects like Neutral Network Simulation & RenderFarm @ Home (to mention a couple) that I can't attach my new hosts to anymore.

The other thing is 4 CPIDs in 1 host and none of them are the same for each project on each host. So when a different host connects to the server it changes the project CPID, then the next host that connects changes it and so on and so forth, day after day.

I'm just grinning and bearing it.
5) Message boards : Number crunching : Credit Calculations. (Message 4381)
Posted 22 Jul 2008 by KAMCOBILL


@ KAMCOBILL:

Hmmm...

I looked over my logs and the event I had was about the time your problem started.

However, I don't use BAM or any other account manager, and my logbook noted that the most likely cause for mine was that Willy was experiencing some growth problems at the time and having some DB difficulties as a result.

So since I have no experience trying to get a CPID to sync up in your situation, I can't really say whether your next attempt strategy is good, bad, or futile! :-(

All I can say is it sounds like a plan, good luck, and I hope it works for you this time! ;-)

Alinator



Willy says he can't do anything about or with CPIDs (just the projects) so that can't be it. Only thing I know is my hosts changes the Project CPIDs at will.

It should work in a day or so after I change all the CPIDs to 1. Maybe less if I update all of them. It'll take some time to detach and change CPIDS though.
6) Message boards : Number crunching : Credit Calculations. (Message 4373)
Posted 22 Jul 2008 by KAMCOBILL


I'm not sure what happen in Nov 2006 besides one of my hosts crashed and then the problem bagan. All my old projects had there CPIDs stolen away in this transition. :D What I can't figure is how can a host have different CPIDs (4 in most of them) and control/change the Project CPIDs when they connect. I thought the Projects Servers controled the hosts. The one good thing is I got down from 10 CPIDs to 4. I even detached from BAM to see if CPID Problem would clear up but it didn't

When BAM retires a project they detach all the hosts from the project and it keeps detaching them for an eternity. An example is ABC@Home Beta. It only has work as needed (Nov 2007 was last WU I finished) & it wasn't reachable for a while so I guess that is why they retired it. I attach to it ever once in a while to see if it up and as soon as my host syncs BAM it detached. If It does give out any work, I'll have to Detach from BAM to run it.





Hmmm...

Now that is interesting. I seem to remember loosing a couple of host CPID's on BOINCStats around then, but everything resync'ed on the same Account CPID.

The only side effect was a couple of hosts lost their graphical data for everything from before the event.

Let me check my logs and get back to you on this.

Alinator



I tried everything to solve the problem. Including begging the Dev (which is a low priorty issue). I would like to see if they could straighten out the problem, but I think I going to detach 15 hosts from all the projects, delete/clear out all BOINC. Then go into one node and change all the CPID to my main one (or suspose to be my main one) then run it until I get all the projects to Sync with my host. Then put the others back online again.
7) Message boards : Number crunching : Credit Calculations. (Message 4370)
Posted 22 Jul 2008 by KAMCOBILL
OK By Gosh I think I got it. Give 1 credit per CPU second for all projects.

So my 200 MHz is worth the same a a Core 2 3GHZ? This does not seem very bright.


Sure it is. Some people are only running 200Mhz because they can't afford to buy one with all the beeps and whistles. If there is going to be Parity why not threat them equal. Actually they should get more credit since they are working so much harder.
8) Message boards : Number crunching : New WU Length? (Message 4363)
Posted 22 Jul 2008 by KAMCOBILL
My Pentium III says "Goodbye" to Milkyway.
The new workunit-length is killing my machine!!!
When Workunit-length is shortened, mayby he ┬┤ll be back ...


Oh My. I have a Pen D 3.0G with 15 MW WUs that the deadline is in 6 hours. That's because of the 9 mins TTC to 9 Hours TTC. They just started running yesterday so I'll only complete 7 of the 20 because of the change of length.

My PII 266Mhz ran one in 68+ hours :-). There shouldn't be anymore flood of WUs once BOINC Manager get the completion right. Nobody will have to complain on only 20 WUs per host now.

Edited: Looks Like I'll have a good many to abort.
9) Message boards : Number crunching : New WU Length? (Message 4359)
Posted 22 Jul 2008 by KAMCOBILL
I'm not sure which projects it was (I thought it was this one but maybe not I'd have to check posts) in early stages the WUs were aborted with because maximum CPU time exceeded. Anyway, I thought it was the projects that was aborting because I don't see a setting In BOINC Manager or BAM to set a CPU Max Time and they weren't running too long before being aborted.

Actually, My PIIs don't do too bad. They are definately better than the Cyrus 100 & AMD 100 I retired. Never tried an AMD 25 though :D. I do have more MW WUs left in the PII that aren't going to make it by deadline because when I got them Time to finish was like 20+ minutes but now it's 68+ hours.

Hehe. My Pen D 3.0Ghz had a hard time meeting the deadlines but they did. It went from 6 to 9 mins to 9+ hours on them. Hopefully, they are ready to crunch with ease again.

The max run time is set by the project administrators. I believe that Orbit recently had this problem.


That's what I tought it was the projests but wasn't sure. I got some of those long WUs OAH come out with too. More that one on some hosts. :-)
10) Message boards : Number crunching : Credit Calculations. (Message 4345)
Posted 22 Jul 2008 by KAMCOBILL


Also, I think there is something funny with the RS project's server code. If RS was the oldest project, it still won't sinc up properly sometimes. Add to that, RS being currently down...


Hmmm...

Well, that might be due to RS having the secondary objective of porting BOINC to php.

OTOH, I haven't had any trouble with my CPID sync'ing up because of RS (or any other project for that matter). Although it did seem to take longer than usual after joining. IIRC, it took about a week or ten days to sync to my 'permanent' CPID instead of the usual couple of days after the first task completed and went through a stat xml cycle.

I hadn't thought about what the effect would be if RS was the oldest project or what would happen if the oldest project wasn't available though.

Of course, I don't have that many hosts or run anywhere near the number of projects available today. The one thing which is true is SAH is my oldest project and all my hosts are attached to it. The other thing which is true is there are still some basic operational areas in BOINC which are kind of shaky and can definitely rear their ugly head given the right set of conditions (like the primary topic in this thread, for example). ;-)

@ KAMCOBILL: Based on your last post (about the CPID problem anyway), this would seem to indicate a BAM problem rather than BOINC per se. I never ran any of the 'retired' projects. However, I used to run Predictor but it remains linked to my CPID even though I haven't crunched a task for them in ages, and have gone months without even doing an update on any of the hosts which are still attached to it.

Alinator


I'm not sure what happen in Nov 2006 besides one of my hosts crashed and then the problem bagan. All my old projects had there CPIDs stolen away in this transition. :D What I can't figure is how can a host have different CPIDs (4 in most of them) and control/change the Project CPIDs when they connect. I thought the Projects Servers controled the hosts. The one good thing is I got down from 10 CPIDs to 4. I even detached from BAM to see if CPID Problem would clear up but it didn't

When BAM retires a project they detach all the hosts from the project and it keeps detaching them for an eternity. An example is ABC@Home Beta. It only has work as needed (Nov 2007 was last WU I finished) & it wasn't reachable for a while so I guess that is why they retired it. I attach to it ever once in a while to see if it up and as soon as my host syncs BAM it detached. If It does give out any work, I'll have to Detach from BAM to run it.

11) Message boards : Number crunching : Credit Calculations. (Message 4333)
Posted 22 Jul 2008 by KAMCOBILL
OK By Gosh I think I got it. Give 1 credit per CPU second for all projects.
12) Message boards : Number crunching : New WU Length? (Message 4322)
Posted 22 Jul 2008 by KAMCOBILL
I'm not sure which projects it was (I thought it was this one but maybe not I'd have to check posts) in early stages the WUs were aborted with because maximum CPU time exceeded. Anyway, I thought it was the projects that was aborting because I don't see a setting In BOINC Manager or BAM to set a CPU Max Time and they weren't running too long before being aborted.

Actually, My PIIs don't do too bad. They are definately better than the Cyrus 100 & AMD 100 I retired. Never tried an AMD 25 though :D. I do have more MW WUs left in the PII that aren't going to make it by deadline because when I got them Time to finish was like 20+ minutes but now it's 68+ hours.

Hehe. My Pen D 3.0Ghz had a hard time meeting the deadlines but they did. It went from 6 to 9 mins to 9+ hours on them. Hopefully, they are ready to crunch with ease again.
13) Message boards : Number crunching : New WU Length? (Message 4318)
Posted 22 Jul 2008 by KAMCOBILL
OK I finished my first WU with PII 266Mhz. It took 68+ CPU Hours. And look Here?:

44228711 44988608 16 Jul 2008 19:23:43 UTC 21 Jul 2008 19:54:27 UTC Over Redundant result Cancelled by server 0.00 --- ---
44228710 44988607 16 Jul 2008 19:23:43 UTC 21 Jul 2008 19:54:27 UTC Over Redundant result Cancelled by server 0.00 --- ---
14) Message boards : Number crunching : Credit Calculations. (Message 4315)
Posted 21 Jul 2008 by KAMCOBILL
That criteria has been meet, plus Paswword, URL, Country, Zipcode, Team and etc. are the same for all projects. All old projects are and have been attached. The only ones I can't keep attached are the one that BAM retired and keep detaching on Sync with BAM.
15) Message boards : Number crunching : Credit Calculations. (Message 4309)
Posted 21 Jul 2008 by KAMCOBILL
Credits Calculations vs. Credit (CPID) Problems? I wish as much effort would go into getting the Credits straight as in the Granting of Credits/cross project parity.

My Credits http://boincstats.com/stats/boinc_user_graph.php?pr=bo&id=b65b36154e3a49f636a3148d4e0b3c83 have been seesawing back and forth since Nov 06. Actually, I'm not into this for Credits, but it would be nice to look at once in a while to see how my efforts compared to others. So as it stands for me, it doesn't matter if I get 1 or 1000 Credits/Hour because it doesn't show in any stats that use CPIDs for their stats.

Since I only have 2 cents I don't want to spend it all here. :D Patience is a Virtue.

You have two different CPIDs. The solution: Attach one machine to ALL projects you have done work for. Cycle through the projects with update giving at least three complete successful cycles.

Second solution. The BOINC client and server have (very recently, not released yet) been modified to incorporate an extra time stamp on the CPID. This will eventually fix the problem automatically.


Hi JM7,

Kinda close. I have 4 CPIDs which have been there since November 2006. I have 16 Hosts attached to 70+ projects (not getting work on all of them). The problem with the CPIDs is in the hosts. Some hosts, actually, have 4 project CPIDs in it. Each time a different host connects to a server it changes the project CPID. If my 4 CPIDs are checked, some of the projects are listed in 2 or more CPIDs (depending on the time of updates) and they switch back and forth.

A fix for my host is what is needed (If I knew how to change all my Projects CPIDs in my hosts to one, I would do it). I know when I look at the Project CPIDs it has a Hash and in most hosts there are 3 to 4 different hash codes.

Theses are the 4 Projects CPIDs in Host PC-13:

HASH(0xe8b240)
HASH(0xe9bb30)
HASH(0xe9bb40)
HASH(0xe8b320)

and here's PC13client_state.xml:

http://www.xgrubberskickass.com/Misc/PC13client_state.xml

All host are like this. The only difference is CPIDs for each project are different from PC-13s

I did a ticket at dev which a fix was projected in 6.2.x and it looks like we're going into 6.3.x shortly. Sure could use a Sync for host CPIDs for the peojects

Hehe, 19 months of this so I guess there's no hurry.

Bill

Edited: Note: Email Hashes are the same.
16) Message boards : Number crunching : Credit Calculations. (Message 4297)
Posted 21 Jul 2008 by KAMCOBILL
Credits Calculations vs. Credit (CPID) Problems? I wish as much effort would go into getting the Credits straight as in the Granting of Credits/cross project parity.

My Credits http://boincstats.com/stats/boinc_user_graph.php?pr=bo&id=b65b36154e3a49f636a3148d4e0b3c83 have been seesawing back and forth since Nov 06. Actually, I'm not into this for Credits, but it would be nice to look at once in a while to see how my efforts compared to others. So as it stands for me, it doesn't matter if I get 1 or 1000 Credits/Hour because it doesn't show in any stats that use CPIDs for their stats.

Since I only have 2 cents I don't want to spend it all here. :D Patience is a Virtue.
17) Message boards : Number crunching : New WU Length? (Message 4159)
Posted 16 Jul 2008 by KAMCOBILL
Sorry for all the confusion, I did make a post about the increased time for WUs when I did it, but I realize that "at the same time" wasn't enough time for some of you guys to realize what was going on. Thanks to those that pointed the others in the right direction.

The answer to the big question is YES, the deadlines will be changed. The only reason it hasn't been done yet is that I do not know how, and Travis is out of town until today or tomorrow. However, it will be fixed ASAP.

As for all the credit issues, my personal belief is that there should be a standard BOINC implementation that gives 1 credit per minute crunched. Simple sweet and no worrying by the project admins if there is too much or too little. But, that's just my opinion. However, I do understand why it's not universal and, so we deal with it.


I'm a little confused here. BOINCManager handles how many work units we get. if we don't need any work we don't get it. Even if the 20 WUs Limit is not reached. Also, the project seem to handle it too. For example:

7/15/2008 15:28:21|Enigma@Home|Message from server: No work sent
7/15/2008 15:28:21|Enigma@Home|Message from server: (won't finish in time) BOINC runs 98.9% of time, computation enabled 99.7% of that


The new crunch time for the WUs are around 7 hours give or take a few. If the deadline is increased, won't they just sit around longer before starting. I have WUs with 30 days deadline and they sit around for 28 days or so. Which also happens to the the MW 7 minute ones. They sit around until the last day of the deadline.

Now of course, I'm one that is attached to 70+ Projects. I don't have enough boxes to run 1 per project, but I'm working on it. I'm crunching with Quads, Dual-Cores and down to PII 266s so I can check how long the new WUs take to finish on the slow boxes. Hehe, I even have a Cyrus 100 And an AMD 100 that I can put Online. Too long of a deadline seems to defeat the turnaround time needed for fresh data.
18) Message boards : Number crunching : Per Host Limit (Message 519)
Posted 21 Nov 2007 by KAMCOBILL
Hello To all fellow crunchers here at milkyway@home

I noticed a while ago in the messages tab:

19/11/2007 15:46:11|Milkyway@home|Message from server: No work sent
19/11/2007 15:46:11|Milkyway@home|Message from server: (reached per-host limit of 8 tasks)

Is this right?
Are there new limits being set?

As previously the quota was set in the thousands per host.

Kind Regards,

John Gray :0)


At the moment there is a limit of 2000 units per core per day. But the project is set to limit the number of units per host to 8 at one time. As these are reported back you receive more. The only people that might have a major problem with this would be those that don't have a continuous connection and run this project exclusively.


11/21/2007 02:09:28|Milkyway@home|Sending scheduler request: To fetch work. Requesting 100470 seconds of work, reporting 0 completed tasks
11/21/2007 02:09:33|Milkyway@home|Scheduler request succeeded: got 0 new tasks
11/21/2007 02:09:33|Milkyway@home|Message from server: No work sent
11/21/2007 02:09:33|Milkyway@home|Message from server: (reached per-host limit of 8 tasks)

oops I forgot. edited:

I only have 1 WU left and still get this message.




Main page · Your account · Message boards


Copyright © 2017 AstroInformatics Group