LHC@home

Work was available for the taking on the server from before 0300 25th to after 0900 26th November 2006. This is the longest we have continuously had work on the server since the summer.

The only reasons I can think of why you might have missed out are

1/ Your box was in 7-day deferral (can only happen on some versions of the client, and only after network problems)

Advice: upgrade your client to 5.4.9 at least. The current version of the client retries every 24hrs after a net problem, so should have picked up work this time

2/ Your box was in "no work fetch" due to other projects' cache sizes throughout that 30 hours.

Advice: halve the cache size, and leave for a few days, repeat halvong the size twice a week till you never see "No work fetch" for more than 2hrs at a time.

Or go straight back to the default 0.1 day setting, which is good for most people unless you are on dial up.

3/ Your box was detached from LHC, or you had suspended the project or set "No new work"

Advice: don't do this. Let the client check in with LHC every so often - it does not cost much for your cpu to do this, and you will get work

4/ Your turn your box off at weekends.

Advice: work quite often turns up here at weekends. If you have to turn the box off, then you have to. But do accept that means missing out on work if any comes up.

5/ You had local network problems throughout the 30 hours.

Advice: get local advice

Edit: added two more to list, no advice for these I'm afraid

6/ Machine died / at menders / having new HD fitted / etc

7/ On holiday and turned boxes off for safety / economy reasons

Edit2, especially for @nne

8/ You have a Mac

Apology: this project has never converted its application to run on a Mac and there are no plans to do so, sorry. Thanks for your interest and good luck with other BOINC projects.

River~~

---

Yes, yes, YES!!!

I was scanning one of my machines when the block of work hit the server. BOINC really does work! With my computers all set to the 0.1 day setting, it took less than two hours before all my machines were happily humming away on LHC WUs.

I can only second the advice that River~~ gives in his posting --> "make sure your client is 5.4.9 or higher", and "let your client check in with LHC every so often". I'm sure that I didn't get as many packets (WUs) as others, but it was enough to make me feel that I had a hand in the project.

To those who might have missed the work for one reason or another, I can only say that you need to retune and reset your client, find another project that will keep your machine(s) busy, and stand ready for the next time -- I am ...

---

If I've lived this long, I've gotta be that old

:-)
I dont know if i should laugh or cry?
There can be another reason :-) I got 30 WU on three machines, thats the best part. The bad part, my daughter inserted a floopy from school and contaminated two machines with W32. Chir of various flavours, which in turn infected amongst other things the WU"s. When i ran the Anti Virus then it of course had to remove the virus, so eight WU"s got destroyed :-( of LHC twenty of other projects?
:-) Dont worry i allowed BOINC to report the corrupted WU"s so they must have been reissued.
Regards
Masud

Work was available for the taking on the server from before 0300 25th to after 0900 26th November 2006. This is the longest we have continuously had work on the server since the summer.

The only reasons I can think of why you might have missed out are

1/ Your box was in 7-day deferral (can only happen on some versions of the client, and only after network problems)

Advice: upgrade your client to 5.4.9 at least. The current version of the client retries every 24hrs after a net problem, so should have picked up work this time

2/ Your box was in "no work fetch" due to other projects' cache sizes throughout that 30 hours.

Advice: halve the cache size, and leave for a few days, repeat halvong the size twice a week till you never see "No work fetch" for more than 2hrs at a time.

Or go straight back to the default 0.1 day setting, which is good for most people unless you are on dial up.

3/ Your box was detached from LHC, or you had suspended the project or set "No new work"

Advice: don't do this. Let the client check in with LHC every so often - it does not cost much for your cpu to do this, and you will get work

4/ Your turn your box off at weekends.

Advice: work quite often turns up here at weekends. If you have to turn the box off, then you have to. But do accept that means missing out on work if any comes up.

5/ You had local network problems throughout the 30 hours.

Advice: get local advice

Edit: added two more to list, no advice for these I'm afraid

6/ Machine died / at menders / having new HD fitted / etc

7/ On holiday and turned boxes off for safety / economy reasons

Edit2, especially for @nne

8/ You have a Mac

Apology: this project has never converted its application to run on a Mac and there are no plans to do so, sorry. Thanks for your interest and good luck with other BOINC projects.

River~~

---

I got WU's... because:

1/ My hosts check frequently for work

2/ All my other projects are on no work fetch while LHC is always allowed to fetch

3/ I never detach from LHC....

4/ I had my box running during the weekend

5/ My ISP is pretty reliable.... no complaints in more than 2 years

6/ none of my hardware died lately

7/ I had my machines on to finis shome QMC work they had... which is now pushed back behind LHC

8/ I have windows only.

I got a nice load of work during those 30 hours.

The work that appeared on 27 November is also
crunching away happily.

---

How sweet it was to wake up Saturday morning with so many work units to crunch!

Thanks River for your reposting the advice by Steve Cressman back in June! :-)

peter

---

Yup, got only 72 WorkUnits (mainly because Caches were filled with other Projects and LHC suspended as usual), but those were my first after 18 Jun 2006 ;)

---

Scientific Network : 45000 MHz - 77824 MB - 1970 GB

Got plenty of work on all my machines.
What is curious is how some machines may have gotten 20-30 wu's where as others received over a 100.
Or 3 machines that are exactly the same and one of them got about 10 times the wu's of the others.
It would make sense if for example, the quad core got 200 wu's and a P4 3.0HT got 20.

I know Einstien tries to send easier wu's to lesser machines and heavier wu's to bigger cpu'd systems, I have found this to be true, ymmv.

Happy to be crunchin !

---

:-) 28 WU"s today, wonder when they came in? becuse as soon as i saw them i connected the machines on dial up and no luck :-(
Regards
Masud

---

:-) 28 WU"s today, wonder when they came in? becuse as soon as i saw them i connected the machines on dial up and no luck :-(
Regards
Masud

The answer to this kind of question can usually be found on Scarecrow's site - he has graphs going back months and numberical data going bach 7 days. Between 1100 and 1500 on 29th we see the work in prograss rising but the results available to send sticking at zero From the graphs I deduce - there were a few thousand WU dribbled in between 1100 and 1500 but that they were put on the server slower than the scheduler could issue them - so it is likely some other people tried to get work between these times and failed.

When Scarecrow's script looked at 1505 they had all gone. You got yours in the last 5mins of availability - this was very lucky

Please post your lotto numbers ;-)

R~~

A look at the pending WorkUnits will reveal the next batch (workunits expiring deadline).

How many will be re-issued remains in question (I assume around ~5%).

Based on my few still pending, I expect them to be re-issued between :
02 Dec 2006 2115Z and 03 Dec 2006 0400Z
(Z=Zulu, equals GMT)

Since I wasn't the first to connect for the last batch and not the last, extend the time window accordingly if you have earlier and later WorkUnit deadlines from your pending list.
It'll be a good idea to have some machines check for work in that Time Window... it should span the same ~30h Window we had when the batch was issued, just alot less in numbers :)

---

Scientific Network : 45000 MHz - 77824 MB - 1970 GB

Hmmm --

Not a bad guess, but off time a bit. Work showed up about 2200 UTC on 3 Dec, and closed around 0015 UTC on 4 Dec, 2006. Probably around 9,000 units ....

---

If I've lived this long, I've gotta be that old

Really a good guess I would say. Just got 18 re-issued work units!
Peter

---

Glad it worked for some :)

Unfortunately my 3 Systems left active on LHC didn't check at the right time and didn't see any work, but that's back to "business as usual" ;)

---

Scientific Network : 45000 MHz - 77824 MB - 1970 GB

:-) Sorry so many forums and message boards? I sometimes wander off:-) Dazed would be the better word.
Thanks, i will look up Scarecrows site. By the way what is lotto number even though i can take a guess at it :-)
Regards
Masud.

:-) 28 WU"s today, wonder when they came in? becuse as soon as i saw them i connected the machines on dial up and no luck :-(
Regards
Masud

The answer to this kind of question can usually be found on Scarecrow's site - he has graphs going back months and numberical data going bach 7 days. Between 1100 and 1500 on 29th we see the work in prograss rising but the results available to send sticking at zero From the graphs I deduce - there were a few thousand WU dribbled in between 1100 and 1500 but that they were put on the server slower than the scheduler could issue them - so it is likely some other people tried to get work between these times and failed.

When Scarecrow's script looked at 1505 they had all gone. You got yours in the last 5mins of availability - this was very lucky

Please post your lotto numbers ;-)

R~~

---

I finally got some.

---

I finally got some.

---

:-) I got 36 WU's :-) Thats the best part, the worst is that one of my machines again destroyed some :-( Is it possible to find out the reason for it to happen, i would be grateful for inputs :-)This is my shiny new P4 Prescott which is doing so, my P3, P2 and 1.7 Gig Celron are happy crunching them? Confusing aint it.
Regards
Masud.

---

Wellll, yes -- it is confusing.

Looked at your lists, and your dual processor machine has completed all the work allocated to it, and you received credit for each WU. The other machines are still crunching, but there's no indication of a problem with the work sent out on the 10th of December.

Only hiccup I see is from the work that was sent out on the 25th of November, and you had several download errors on one machine, but that machine behaved properly this time -- might have been a random problem with your server

---

If I've lived this long, I've gotta be that old

:-) Sorry i got confused. The reason was, that this lot were WU's of less then ten seconds :-( This is the first time i have ever come across those. Thanks to heavens i did not open up the machine and format it. PheeeW! that would have been a hassle :-) Lucky me ;-)
Anyway. I am happy that for the last two weeks every few days fifteen WU's some how dribble into my machine :-) Usually after midnight GMT :-) when nobody is around :-) Otherwise i would have grabbed a handful more :-(
Some how, my "To Completion Time" and "Completed Time" are nowhere same. Is it possible to explain the reason?
Regards
Masud.

---