LHC@home

I have come to the conclusion that the interest in the LHC Project have waned greatly due to the 0 Granted Credit, Lack of Consistent WU's to do & other various Problems.

After looking at some of the Stats Pages & going through some of my Pending Credits that are approaching 2 weeks old already save but for a handful of people nobody's really Crunching for LHC anymore...The Stats Pages show me that out of the 6000 Clients or Hosts that only about 10% - 15% of them are actually returning any results ... :(

I found Account after Account where people have been sent WU's but their not Crunching them, and the same people haven't Crunched any for over a month now, yet they get new WU's that just simply sit there and reach their Deadlines & then have to be sent out again ...

I sadly think this Project has gone the way of the Predictor Project and theres just not enough interest in it anymore ...

> I sadly think this Project has gone the way of the Predictor Project and
> theres just not enough interest in it anymore ...
>
This sounds really sad. I hope you're not right... I'm sure that when LHC goes full throttle, people will return. The issue of not returned WU's is a problem, though!

sadly,

sysfried

---

I doubt it. There has not been work for most people. It seems like they get impatient and quit instead of just letting BOINC do it's job and get back to the project later.

---

BOINC WIKI

BOINCing since 2002/12/8

There has not been work for most people.
==========

Almost every Account I looked at the Client had new WU's to do, some of them as many as 25-30 new WU's...But most of them where due by the 22 of April so you know their not going to get them done in time, especially since they haven't crunched anything in the last month ...

I'm sure the LHC Project will always have it's hardcore clients that will continue to crunch diligently, but if the Project can't attract and keep a larger Client-el Base thats willing to crunch the WU's then they are going to have a hard row to hoe I think in getting the WU's done ...

PS: I went to the Einstein Project & the 1000 Client has a 328.42 RAC where here at the LHC Site the 1000 Client has 10.97 RAC, that should tell you how much more interest there is in the Einstein Project compared to the LHC Project ...

> But most of them where due by the 22 of April so
> you know their not going to get them done in time, especially since they
> haven't crunched anything in the last month ...
>
@ PoorBoy
IMO:As they were send out on 4.-9. of this month: they are all 4.64 (which caused such a lot problems)!!!
So it seems : Those users have restted the project or detached from the project, when it was clear; that they risque to waste CPU-time, when they crunch 4.64.

I hope those WU will not be resend with executable 4.64!!!
(with 4.66 and higher, they are 'crunchable')

That is only a try of explaination
littleBouncer

I hope those WU will not be resend with executable 4.64!!!
(with 4.66 and higher, they are 'crunchable')
=========

They more than probably will be sent back out as 4.64's, but 1 can hope not ... What confuses me is some of the WU's I have already be Granted 0 Credit for even though I successfully returned the Results in time & they were Valid WU's are being sent back out again ...

For all I know I can receive the same WU that I received 0 Credit for on another 1 of my Computers and I get the dubious distinction of Crunching it again for nothing ... D'oh

What is the purpose of making somebody else crunch the WU again when there is no chance of them receiving any Credit either ... These are some of the things that drive people away from the Project ...

The problem is that boinc does not bind a WorkUnit to a specific version of the executable. When we released the new WUs some people simply kept the old v. 4.64 sixtrack and used that to compute with, even though the 4.66 was on the server. I do not know why the boinc client does not automatically download the newest version of the executable.

When this happens all we can do is encourage people to reset the project. When people reset they get the newest version.

This means that when a WU is re-issued whatever version is on the machine it is issued to is used.

---

Chrulle
Research Assistant & Ex-LHC@home developer
Niels Bohr Institute

When this happens all we can do is encourage people to reset the project. When people reset they get the newest version.

This means that when a WU is re-issued whatever version is on the machine it is issued to is used.
=========

Thats what I did (Reset the Project) after seeing WU after WU of the 4.64 reset back to 0:00 time when they finished. So I reset any computers that had the 4.64 on them and got the newer Versions downloaded, but we still recieve 0 Credit for some of them because if somebody didn't reset and showed 0:00 Time then everybody else still has to go by that too ...

> The problem is that boinc does not bind a WorkUnit to a specific version of
> the executable. When we released the new WUs some people simply kept the old
> v. 4.64 sixtrack and used that to compute with, even though the 4.66 was on
> the server. I do not know why the boinc client does not automatically download
> the newest version of the executable.
>
> When this happens all we can do is encourage people to reset the project. When
> people reset they get the newest version.
>
> This means that when a WU is re-issued whatever version is on the machine it
> is issued to is used.

I have spotted a bug (during the closed beta of F@h) that if the person aborts a WU via gui and the new WU has a later application version, the new science applciation is downloaded but the current (older application is used.

The bug is reported at http://bbugs.axpr.net/bug.php?op=show&bugid=202

If you need any additional alpha people drop me an email.

---

73 de Peter VK3AVE

If you need any additional alpha people drop me an email.
========

I tried to create an Account over at the Alpha Site too, but it's disabled for now ...

This project is still somewhat in the testing stage of work. I think as time goes on there will be more work and probably more Science Applications. I mean, I also wish for more work. But we just have to take what we can get and smile.

If there is the possiblity that work will not be finished fast enough they can open up the sign-ups again and increase the Participant base at any time ...

As this project is a major interest area for me, well, they will have to shoot me to get me to leave.

I even suspect that this may become a significant project class for BOINC with other accellerators (perhaps) writing Science Applications to model the expiriments before comitting accellerator time to an expiriment ...

One of my earliest memories of this project was the expected completion time ... which we "beat" by a significant margin ... and this was with fewer participants than we have now.

---

Like it's mentioned in other places, a reset is without any doupt
a helpfull action, the user can take.

While the project Queue on the affected client is empty, why not.

IS there still work sitting on the client,

I would rather do following :

stop/exit the boinc client,

delete the "application" *.exe file(s), they are at defined locations on customers client side.

restart the boinc client

Might the actual processing/processed Work is suffering and returning "bad"

the next work will start and put an internal request for selfdownloading the MISSING files, this time the "applicaion".

This way, a forced "redownload" can be scheduled.

> When this happens all we can do is encourage people to reset the project. When
> people reset they get the newest version.

Ahh, so that's why :D I did see a mention of resetting but I thought the new version would eventually be downloaded automatically - I would just have to be patient.

---

> show me that out of the 6000 Clients or Hosts that only about 10% - 15% of
> them are actually returning any results ... :(

look back in my "results", from 9.4 to now, i saw that the most WU's count, aka are not pending. In this view, i think there must be much more WU that have been processed than Units that still are waiting for the deadline.
But from the "Bad"-Client-WU (4.64) the most of them did goes to dealine.

...

---

> As this project is a major interest area for me, well, they will have to shoot me to get me to leave. (Paul)

I second that. My machines are sitting here with the 4.67 exe, waiting for new work. Of course, they are busy with other BOINC projects while they wait, as is the intent of BOINC. So they will crunch the work when it becomes available, and will not run out of patience. They are computers, after all...

----edit----

> The problem is that boinc does not bind a WorkUnit to a specific version of the executable. (Chrulle)

??? Is this really true? I'm (almost 100%) positive that on other BOINC projects that each WU specifies which EXE it needs to run. Isn't that what the "application" column is showing?

---

It's REAL simple.

<B>EITHER</B>

a) Your dedicated to the Science and will stick with the project until it's cancelled forever, understanding that LHC is a "Work In Progress" and adapting to the sporadic nature of the work as the LHC Project continues to grow and refine the Science and it's programming to fix all the issues wrong with it....

OR

b) You are impatient (for many different reasons) with LHC because you think that they are not resolving what you think are important issues quickly enough, or that they don't take you seriously when you TRY to tell the project admins about serious bugs, client problems, etc...etc... so you decide to stop crunching and move to another project that you consider to be "more stable".

If this were a poll, I'd be voting for <B><I>a)</B></I> as I truly beleive that this particular project has some real potential to revolutionize an important part of the Physics Scientific Community.....

Some of you may think that my sentiment is as silly as those people who look for aliens with radio signals, but I personally think that every BOINC project has some measure of merit and or potential in it's particular field...

So, either stay here with LHC and show your dedication to the science as the problems are all eventually fixed, or leave now so there's more work available for those will not sit with machines filled with work that goes unused because their clients are either too outdated or BOINC's not run often enough to crunch the work before it's deadline passes.

The Choice is your and your's alone to make, but no matter WHAT you decide to think about LHC and the way they do things, <B><I>PLEASE</B></I> do not continue to take work and then not crunch it just because some problem with the client or whatever makes you mad and say "The hell with this project" because that's not FAIR to the rest of us who are willing to work within the peculiar way that LHC currently uses to generate work for our machines to have to do without because you are mad and dont want to give it your cycles.

---

Hi All,
Poorboy, I read what you posted and I went back and checked my pending credits and my results where there in lies some issues. I have the following "pending results" that will time out on April 23, but I do not have any units to crunch, and have not gotten any in a while. I'm not sure if people are waiting on these results or not.
They are as follows:
407306 87240 9 Apr 2005 11:08:35 UTC 23 Apr 2005 8:28:35 UTC In Progress Unknown New --- --- ---
407270 87233 9 Apr 2005 11:02:15 UTC 23 Apr 2005 8:22:15 UTC In Progress Unknown New --- --- ---
407241 87227 9 Apr 2005 11:08:35 UTC 23 Apr 2005 8:28:35 UTC In Progress Unknown New --- --- ---
407213 87222 9 Apr 2005 11:08:35 UTC 23 Apr 2005 8:28:35 UTC In Progress Unknown New --- --- ---
407212 87221 9 Apr 2005 11:08:35 UTC 23 Apr 2005 8:28:35 UTC In Progress Unknown New --- --- ---
407204 87220 9 Apr 2005 11:08:35 UTC 23 Apr 2005 8:28:35 UTC In Progress Unknown New --- --- ---
407202 87219 9 Apr 2005 11:08:35 UTC 23 Apr 2005 8:28:35 UTC In Progress Unknown New --- --- ---
407183 87216 9 Apr 2005 11:02:15 UTC 23 Apr 2005 8:22:15 UTC In Progress Unknown New --- --- ---
407179 87215 9 Apr 2005 11:02:15 UTC 23 Apr 2005 8:22:15 UTC In Progress Unknown New --- --- ---
407175 87214 9 Apr 2005 11:02:15 UTC 23 Apr 2005 8:22:15 UTC In Progress Unknown New --- --- ---
407166 87212 9 Apr 2005 11:02:15 UTC 23 Apr 2005 8:22:15 UTC In Progress Unknown New --- --- ---
407156 87210 9 Apr 2005 11:02:15 UTC 23 Apr 2005 8:22:15 UTC In Progress Unknown New --- --- ---
407091 87197 9 Apr 2005 10:39:06 UTC 23 Apr 2005 7:59:06 UTC In Progress Unknown New --- --- ---
407087 87196 9 Apr 2005 10:39:06 UTC 23 Apr 2005 7:59:06 UTC In Progress Unknown New --- --- ---
407017 87182 9 Apr 2005 10:39:06 UTC 23 Apr 2005 7:59:06 UTC In Progress Unknown New --- --- ---
406988 87177 9 Apr 2005 10:39:06 UTC 23 Apr 2005 7:59:06 UTC In Progress Unknown New --- --- ---
406986 87176 9 Apr 2005 10:39:06 UTC 23 Apr 2005 7:59:06 UTC In Progress Unknown New --- --- ---
406940 87167 9 Apr 2005 10:39:06 UTC 23 Apr 2005 7:59:06 UTC In Progress Unknown New --- --- ---


Respectfully,

Rocky

---

<img src="http://boincwapstats.sourceforge.net/summary.php?name=ML%20Cudd&amp;team=US%20Navy&amp;seti=123863&amp;einstein=4173&amp;lhc=1091&amp;climate=2545&amp;predictor=7916&amp;style=6&amp;ctx=yellow">
WinXP SP2 Boinc - 4.44
Have A Great Day!

As long we have host's like this.....

there will never be enough work for those who realy process them!!!
( over 700 results for 600 credits ?!?)
( OK, this one had 'only' download-errors, but I saw some with as much stated with "client error" )

greetz from Switzerland (and still crunching LHC)
littleBouncer

---

> <B><I>PLEASE</B></I> do not continue to take work and then not crunch it just
> because some problem with the client or whatever makes you mad and say
> "The hell with this project"

I don't think that the same peoples that think so, really read posts here ...
So what does it change..

@mlcudd
Perhaps you did the "project reset" after you saw that you have 4.64 Units?
If so, you have lost this work units.

---

> ??? Is this really true? I'm (almost 100%) positive that on other BOINC
> projects that each WU specifies which EXE it needs to run. Isn't that what
> the "application" column is showing?

Correct, the 'version_num' Linie say what application (version) is used to run a unique unit.

At one WU i try to edit the current_state.xml, and chance the application setting of a single 4.64 unit to 4.66, and after restart it use the 4.66.
By edit the (right) line (by the work unit) you can run a unit with a newer application, BUT it ONLY works if you HAVE allready downloaded the newer application!!! I you don't have it locally (and an entry in the current_state.xml) you LOST the Unit. Be extremly carefully with that file.

(workunit)
(name)v64boince6ib1-XYs4_1234_0_sixvf_1234(/name)
(app_name)sixtrack(/app_name)
(version_num)467(/version_num)
.....
(/workunit)

Did you see the version_num line?.. yes, this is the version what application version is used to run the WU.

But a warning at the beginning.. :-) If you edit ANY LINE in current_state.xml, you can lost ALL work of that client !!!!!!!!!
This is the most important file!!!

---