21) Message boards : Number crunching : only 10 k WU's left (Message 9213)
Posted 8 Aug 2005 by Jayargh
Post:
> As I see it LHC is a project that needs fast turn around and not long local
> queues for people obsesed with credit after everybody else run dry.
>
> So ideal cruncher is the one that downloads one WU, proces it, return it and
> then dowload another. OK, perhaps three at a time. There is no use that
> someone sends results back after 10 days and after 4 have procesed it. Some
> users have set connect to 10 days and do multiple projects.
>
> So, I believe that server should be set in the way nobody gets more than 0,5
> days at the time. It could be also set to initialy send to only three people
> and quorum of two is enough as identical results must be produced. With
> deadline like 5 days or even less, throughput will dramatically improve.
>
> And for those who only think about credit. We are doing this the way program
> manager and science requires. If you don't like it, there are many other
> projects you can devote your CPU time.
>
> Tony
Tony,
Would like to say in all due respect you are seemingly extremist in your view. 1st off if you care about the "science" did you ever think that maybe a 3 quorum also gives the scientists more accuracy ( or less sometimes) making them question the validity of certain results? Also the 2 results that may come in after the quorum might shed more light on a wu's problem. I keep a 3 day cache because I believe it is prudent given my situation. Using BOINC 4.72 is going to allow me to bring my cache close to 3 days and be able to allow more work to other projects(because times no longer skewed) I am not a hoarder as you would make out and my results I insist are just as valid and important to the LHC team as yours ( some of us don't also have the TIME to edit granted wu's out of our work that is a luxury even if I agreed with your philosiphy)
22) Message boards : Number crunching : Thanks 2 LHC Admins (Message 9212)
Posted 8 Aug 2005 by Jayargh
Post:
> Since we're running low on work again (I hope this changes tomorrow), I'd like
> to say a big "THANK YOU" to the LHC Admins for providing us with work without
> blackouts for 30 days straight now! I LOVE IT! :-D
>
> Cheers to:
> Markku (I know you're with us somehow)
> Chrulle (You're just great!)
> and of course the other LHC admins...
>
> Sincerely,
>
> Sysfried
Yes Sysfried agreed,(good post-guess this means we really made it to production mode)and I would like to add Ben Segal's contribution to admin as a front man :)
23) Message boards : Number crunching : only 10 k WU's left (Message 9211)
Posted 8 Aug 2005 by Jayargh
Post:
> Our problem is exactly turnaround. We cannot do post analysis until we get all
> the WU in a study back. This may take as much as 6 weeks with a 14 day
> deadline. The idea with the automatic deadline estimater is that it will lower
> the deadline as long as it will not cause too big a dent in the throughput.
1st off Chrulle would like to say I'm sorry if my last post was seeming like jumping on you for "perceiving" you were changing 2 week deadline..... As I stated I believe your "Turnaround" is taking you tooooo long... dont think resends quicker would hurt more than 5% production. What Chrulle is ideal turnaround (realisticly not ideally) if 6 weeks seemingly too long?
24) Message boards : Team invites : 2CPU.com Team Invite (Message 9205)
Posted 7 Aug 2005 by Jayargh
Post:
We have a policy of open enrollment all are invited with no restrictions.
25) Message boards : Number crunching : This guy really needs to check his computer, or overclock less... (Message 9194)
Posted 6 Aug 2005 by Jayargh
Post:
> > > With this one, the problem is, BURP somehow took over LHC. (or was
> it
> > the
> > > other way around?)
It was Burp my best guess as this is NOT the 1st time and is why Burp has announced major issues with the developers of BOINC.... I was fortunate as I am sure many were because I stayed attached and because the abrupt downtime of Burp for sooo long my client was at like 8 days to connect so I punished Burp and passed the corrupt wu's as I let the time to connect ride :)
26) Message boards : Number crunching : In The Works Deadline Change (Message 9193)
Posted 6 Aug 2005 by Jayargh
Post:
Reposting my reply to Chrulles reply due to thread excessively long and really off this topic:Posted: 6 Aug 2005 0:23:09 UTC [Edit this post]
Last modified: 6 Aug 2005 1:39:34 UTC

Currently at being attached (99)% or so to your project, I am set at a 10 day cache with BOINC 4.45 and get roughly 3 days of work(slightly improved since some adjustment on your part of run times.)I am pretty sure that a reduction in deadline would do ME no harm,however others have other circumstances and AGAIN it is currently RIDICULOUS that I set at 10 days and get 3. Please further solve this situation BEFORE you change things.(deadlines)[EDIT] Would like to add if you need more efficiency rather than jack with the 2 week deadline try sending out those results getting no returns more OFTEN than shortening the deadline. Have had a bunch in (recent)past stuck with like 3 unsent workunits for up to 5-7 DAYS!Bang units resent and results received! Chrulle I would also like to say that many MANY people are attracted to the 2 week deadline.... don't be short-sighted by what a program tells you....only....and exactly... people are still people... Perhaps you need a psych major on your staff :)CPDN is a prime example of credit being taken lightly.....thier downfall was forewarned.... You have been successful so far...please keep it that way as I LOVE helping design a super-collider:)[EDIT]


ID: 9188 / Rating: 0 - rate: + / - [Reply to this post]

Chrulle Joined: Jul 27, 2004
Posts: 52
ID: 6
Posted: 5 Aug 2005 23:14:11 UTC

Our problem is exactly turnaround. We cannot do post analysis until we get all the WU in a study back. This may take as much as 6 weeks with a 14 day deadline. The idea with the automatic deadline estimater is that it will lower the deadline as long as it will not cause too big a dent in the throughput.

Chrulle
LHC@home developer

ID: 9187 / Rating: 0 - rate: + / - [Reply to this post]

27) Message boards : Number crunching : only 10 k WU's left (Message 9188)
Posted 6 Aug 2005 by Jayargh
Post:
Currently at being attached (99)% or so to your project, I am set at a 10 day cache with BOINC 4.45 and get roughly 3 days of work(slightly improved since some adjustment on your part of run times.)I am pretty sure that a reduction in deadline would do ME no harm,however others have other circumstances and AGAIN it is currently RIDICULOUS that I set at 10 days and get 3. Please further solve this situation BEFORE you change things.(deadlines)[EDIT] Would like to add if you need more efficiency rather than jack with the 2 week deadline try sending out those results getting no returns more OFTEN than shortening the deadline. Have had a bunch in past stuck with like 3 unsent workunits for up to 5 DAYS![EDIT]
28) Message boards : Number crunching : BOINC Version 4.72 (Message 9186)
Posted 5 Aug 2005 by Jayargh
Post:
Can someone please give me a link to said version as I have heard enough positive to try and BOINC website is not listing it :( It seems 4.45 does well with 1 project or 4+ projects but the tweeners with large diversities in resources cause me to do too much tweaking :)
29) Message boards : Number crunching : only 10 k WU's left (Message 8990)
Posted 28 Jul 2005 by Jayargh
Post:
> >
> The only fix is to stop and restart BOINC. If you are desperate, you can open
> the client_state.xml file and hand edit the LT debts (while BOINC is halted).
>
Sorry JMVII but that is not the only fix.... the short of it for a temporary quick fix suspend all other workunits and/or projects and LHC will download... stopping and starting Boinc does not work...the other part of a longer term fix is changing resource settings
30) Message boards : Number crunching : This guy really needs to check his computer, or overclock less... (Message 8956)
Posted 27 Jul 2005 by Jayargh
Post:
>
> > Does LHC have a download quota? I can't find it anywhere.
> >
>
> Yes - it's set to 100 units per day, I think. It may have changed - at one
> point it was up and down like a fiddler's elbow!
>
> Giskard
>
Yes it is at 100 units a day and comes into serious play top 100 hosts with 100k workunits unlike the million turn we are doing now. Less than an hour apiece (info for the newcomers)
31) Message boards : Number crunching : @Markku - Estimated WU Completion (Message 8905)
Posted 25 Jul 2005 by Jayargh
Post:
> > Be nice if Markku could reply.....
> >
> > And I'm yet to be shown a computer that actually takes longer than 50% of
> the
> > estimated wu completion time (even though both JM7 and Markku have
> indicated
> > that they exist), especially in the latest few batches of wu's.
> >
> > Live long and crunch.
> >
> > Paul.
> >
> I have one that I believe takes around 80% of the estimated time.
>
And yet I have a host that does an average of 23% of estimate....(hence my cache is screwed, have 1.5 days of work and Boinc setting is 10 days...Whats wrong with this picture?).... I hope they release Boinc 4.7x soon(due to Boinc actually supposed to adjust time to completion on ready to run workunits) because it doesn't look like admin will adjust soon.... Chrulle was supossed to look at this 2 months ago according to a response by Markku at same time frame :(
32) Message boards : Number crunching : Where is everyone (Message 8892)
Posted 24 Jul 2005 by Jayargh
Post:
Near Galveston Island ..South Texas... Hurricane capital of the Gulf of Mexico and lately feels like Earths Heat Pump... maybe a few wayward protons from sixtrack are colliding over us and heating up the atmosphere hehe
33) Message boards : Number crunching : Strange! (Message 8852)
Posted 23 Jul 2005 by Jayargh
Post:
At any rate it
> is now a rare problem on the projects that it was a big problem on before.
>
Yes it was paticularly bad over at Einstein with ghost units..... and this sending out 1 unit then more in a day or 2 has been a common experience since LHC got the new server,I have been observing this but never made mention , didn't think it was important enough... guess with more people crunching more things get noticed and posted.....@ Colt... LHC had a 3 validate system and quorum at 2 but this got out of control when 0 was reported by some users due to a problem with sixtrack program and many paticipants got 0 credit when crunching a 100 credit workunit,remember?.... sometimes a good reason to over validate... Beta was a lot of fun...many of us lost a LOT of Boinc credit but thats not my primary concern or the others who have crunched from the beginning. The science is the X factor :)
34) Message boards : Number crunching : The Great LHC Hold-Up (Message 8196)
Posted 28 Jun 2005 by Jayargh
Post:
I believe we have been waiting these 2 weeks for new work when sooooo much was promised because of the "0" problem on different platforms.... but maybe I was wrong.... Twould be a nice gesture from the physicists to explain to us laymen what happened in a few short words even. As being a general partner in this collabaration a bit of info is always appreciated :) [Edit] And oh thank-you much Markku for the update on the when.
35) Message boards : Number crunching : New version of Boinc (Message 8090)
Posted 16 Jun 2005 by Jayargh
Post:
Ok Now I have tried to increase my time to connect from 3 to 4 days instead of changing resources.... on my p4 single thread I go into deadline mode with predictor... says I won't finish b4 next connect time( Don"t have this host set up for posting so no logs sorry) Deadline is 6-21 whats wrong with this scenario? Again a rushed development version made mainstream.... I'm considering setting all back to 4.19 (even with Burp) until they address these issues edit.. guess I need to try 4.19 to see if it is worth giving up all the extra choices 4.45 delivers.Maybe 4.25 is next choice.... this is getting to be tooooo much work with 4.45 1st too much work now not enough...I bet devs gettin an earfull hehe edit
36) Message boards : Number crunching : New version of Boinc (Message 8077)
Posted 15 Jun 2005 by Jayargh
Post:
I agree lb,although I like the options of 4.45 , I have had a host sitting there with no work requesting no work because (I beleive) my resource share for LHC was too high even though I had no LHC work. That is silly and makes you wonder why they don't still call this a development version. It would be nice for once to have Boinc work as advertised and not have to manually check hosts all the time. At least 4.45 is not in constant panic mode doing edm like 43 and 44 were and honors resource settings
37) Message boards : Number crunching : first bunch done... now what? (Message 8071)
Posted 14 Jun 2005 by Jayargh
Post:
> >
> Well, in the past I did not join teams as I was not wanting to, ahem,
> advertise for one over the other. Nor seem to endorse any.
>
I'm very sorry for you that your "position" keeps you away from the fun of being on a team. I mean Boinc is many things to many people ...its not ALL about the science and or stature of the individual ...but it IS about individual choice which EVERYONE should have the right to excercise and Paul it IS fun being on a team... I went seti teamless in classic since I had joined in 2000 and round about it was cpdn got me here in Boinc and being on teams have made it a much more enjoyable experience ( whether its my team or not)
38) Message boards : Number crunching : first bunch done... now what? (Message 8068)
Posted 14 Jun 2005 by Jayargh
Post:
and on a side note Paul... I urge you to check out 2CPU.com's many informative articles and users on the subject of optimizing yourself :) maybe we can help you enough ,you join our team (wink wink)
39) Message boards : Number crunching : first bunch done... now what? (Message 8066)
Posted 14 Jun 2005 by Jayargh
Post:
>
>
> Yes this is very right, at seti you get the most return of investment for the
> cycles spend.
CPDN is by far the highest...look who is highest in total boincstat numbers its most cpdn some seti mixed in...they went from 75 credits per trickle to 95 now far exceeding all other projects ... believe the consistency of the #s allows it to outproduce seti by at least 3-5%
> With supporting several project the theoretical speed is also cutted down.
> Thats might so much people are just taking 1 project and crunching like a *ell
> for it.
> This is why I love LHC I now devote 100% when we have work but can easily switch and support other projects in downtime...another personal "challenge" to do this efficiently hence the fun in juggling boinc if not for projects (which if for me anyway) but for the fun of the challenge
>
> And it will be summer, outside the bird and more are singing, people are used
> to wear shorter cloth and spend more time outside.
> Well... guess it depends where ya are ric... here in south Texas outside is 100F+ everyday (at least the heat index if not close in temps) The birds are dying not singing lol and no one is outside... they all stay in in thier comfortably cooled homes (as long as you can afford the juice), along with thier puters cooled very sufficiently (wink wink)
> It's to hot inside with all the cpus running...
>
>
>
40) Message boards : Number crunching : first bunch done... now what? (Message 7992)
Posted 7 Jun 2005 by Jayargh
Post:
Are we moving forward slowly because................ From Chrulle a few days ago while we were in delay from a june 1 start.........Well, once again we are stumped.

On the XP sp1 and XP sp2 and 2000 sp 4. The results we get back are strange. The numerical value of the results are the same, but while we on other platforms the representation of 0 is 15 digits and a E on the afore mentioned platforms we get 16 digits and a e. This is not really a problem because the value is the same just the representation is different. So we simply have to change our post processing checks to actually parse the results instead of just running diff.

We are worried though, because we do not understand why this is happening, and although we have not seen it affecting any of the real numbers yet it might still happen. We are doing static linking so there should be no reason for this to happen.


Chrulle
LHC@home developer

This frightens me because noone has come forward to say this issue is resolved. Are we still in a test phase hence the vaunted million workunits so slow in coming?


Previous 20 · Next 20


©2024 CERN