1) Message boards : Number crunching : How often does LHC shut down? (Message 19582)
Posted 30 Apr 2008 by Profile [B^S] Astral Walker
Post:
If CERN doesn't fund LHC, who does? QMC?
If CERN doesn't fund LHC, how come they get to tell you to take the forums down for 10 days (which messed up several thousand BOINC installs)?

I assume it's the university of london that controls the project and cern that controls the content. Or something like that.
2) Message boards : Number crunching : Please note: this project rarely has work (Message 19418)
Posted 14 Apr 2008 by Profile [B^S] Astral Walker
Post:
I suppose it is like their WORLD Series which is only competed for by US teams.

That's completely wrong. Toronto is in Canada and competes as does Montreal which invaded and captured Washington DC and renamed their team.

Of course competitions like the champions league are properly named since obviously only champions are invited. Oh wait...

;)
3) Message boards : Number crunching : shut down for maintenance ? (Message 19384)
Posted 12 Apr 2008 by Profile [B^S] Astral Walker
Post:
Yeah it's not like the WUs can be reissued now anyway so even if this goes on so long that they expired we should still get credit. I have a bunch of WUs on several machines ready to report as well but the deadlines are still almost a week away.
4) Message boards : Number crunching : Initial Replication (Message 19274)
Posted 19 Mar 2008 by Profile [B^S] Astral Walker
Post:
The newer BOINC server software has a solution for that - outcome 221, 'Redundant result - cancelled by server'. It would work particularly well at LHC, because most hosts are contacting the server every 15 minutes while there's work around, and in general the late returns are high cache/low resource share, rather than extended crunching times (at least, for the 100,000 turn WUs they are).

The problem with that is once the WU starts on your system, even for a few seconds, it will not be aborted. At least that's how I understand it.

The length you select for Rosetta WUs essentially determines the number of steps to be executed for particular proteins. This isn't much different than the number of turns in an LHC WU I suppose. With Rosetta, 3 100 step WUs is the same as 1 300 step WU so the number of WUs being calculated, or not being calculated, is not really relevant.

Anyway, I've never been particularly concerned about the IR>Q issue. While I understand that IR>Q has a waste component, wasted cycles are just the byproduct of quality control. Who's to say that there aren't other forms of redundancy (i.e. wasted cycles) built into the WUs of this or other projects? I'm not going to worry about it for now.

I am, though, one of those people that gets irritated by short deadline projects. Resource shares do not work well in the long run, and are even worse in the short term, at least for me, as I change my shares fairly often. The team I'm on has a project of the month and there are other competitions with projects I like such as the recent challenge at Prime Grid and the current one at WCG, and I use resource shares to try and get my machine to crunch those projects with higher priority.

But LHC is always #1 with me. :)

My main computer only connects to the internet sporadically and BOINC becomes very uncaring about resource shares as it will seemingly randomly select the first project it contacts with a share of 1 (LHC is 1000) and then proceed to download 50 WUs. Then it will contact Malaria and download another block of WUs with a short deadline so that the other WUs can't finish on time as BOINC still can't figure out how long a WU will take after months and years of running the project. It's nice to say how BOINC is supposed to work but in this area it does not run as advertised.

So as I get off my soapbox my conclusion is that the real problems with wasted cycles have more to do with how BOINC works than the individual projects.
5) Message boards : LHC@home Science : Amendment of 'Resource share' (Message 19256)
Posted 17 Mar 2008 by Profile [B^S] Astral Walker
Post:
I've found it a bit hit or miss whether different projects accept resource share instructions from BAM/Boincstats some do and some don't (and I suspect other account managers are the same) and it takes a day or so to find out either way. going direct to the projects' your account/project preferences page gives certainty that it'll take.
it's a slight pity as it would be nice to do it seemlessly from a single screen over a number of projects.

The only project that doesn't work well for me through BAM is WCG. Everything else appears to work quite well and instantly.
6) Message boards : LHC@home Science : why are you going to slaughter us all, scientists? (Message 19255)
Posted 17 Mar 2008 by Profile [B^S] Astral Walker
Post:
We don't see a dome of light because our eyes suck and can't see stars dimmer than about 6th magnitude under the darkest, clearest skies.

Anyway, here is why the LHC is the wrong place to look for black holes. They may be produced in my backyard. ;)
7) Message boards : Number crunching : Oldest molding/untouched credit awards (Message 19199)
Posted 10 Mar 2008 by Profile [B^S] Astral Walker
Post:
Here is my oldest WU 1368128. It's from May 20, 2006 and someone still has pending credit from it (not me thankfully).

The host itself is still running but I've changed its guts at least once since then.
8) Message boards : Number crunching : BOINC 5.10.28 (Message 18593)
Posted 13 Nov 2007 by Profile [B^S] Astral Walker
Post:
Thanks.
9) Message boards : Number crunching : BOINC 5.10.28 (Message 18585)
Posted 12 Nov 2007 by Profile [B^S] Astral Walker
Post:
Would that be this change?

- client: allow up to a day (rather than work_buf_min()) to elapsed between completing a result and reporting it.


I wasn't a fan of it either, but apparently the project server take a bigger hit when people return results one at a time, which is what the old reporting policy caused. I can see why they would try to cut down on such things.

I have a high speed quad core machine that I connect to the internet maybe twice a week (until I get a wireless network setup at least) so I usually upload and download a lot of work durnig those times. As I'm running 5.10.20 is that going to effect me?
10) Message boards : Number crunching : work units?? (Message 18211)
Posted 16 Oct 2007 by Profile [B^S] Astral Walker
Post:
Currently there is computing power needed at SIMAP: http://boinc.bio.wzw.tum.de/boincsimap/

as well as FightAIDS@home:
http://www.worldcommunitygrid.org/projects_showcase/viewFaahResearch.do

and Einstein@Home:
http://einstein.phys.uwm.edu/

So who has idle machines could be very useful contributing to these projects.

I've been attached to those projects for a while. In fact I just attached my newly build quadcore but it is choking on all those 40+ hour Einstein WUs it downloaded. :p
11) Message boards : Number crunching : Initial Replication (Message 17703)
Posted 4 Aug 2007 by Profile [B^S] Astral Walker
Post:
[quote]:-) if admin thinks its is important then they will implement your ideas ;-) but why should they go to all that hassle if their requirements are being met? B-)

To demonstrate that they are responsible adults rather than selfish, wasteful sloths, just for openers.

Thanks for the article but if you want it to carry any weight then you need to quote the source. Anyway, I don't see anything in the article that says LHC needs initial replication of 5. Do you?

I don't see anything in the article that says they can't do their research into DC without wasting our resources. Do you?

If you look at any large construction project (which is what the LHC is), the need to meet deadlines always requires added resources. And accuracy naturally requires replications for the reasons stated above mahy times.

Anyway, I don't know the specific article he quoted but the basic idea with the discrepancy between Intel and Intel compatible CPUs is repeated here, which is a paper written by a CERN scientist and some guy from Canada. While they do say they fixed this particular issue, what specifically are your reasons for saying that 3 replications is enough and 5 is just waste? Why is it that if we disagree with you we must be selfish wasteful sloths?

That article also says a replication of at least 3 is required for accuracy. I guess those folks at CERN (edit: and Canada) are selfish, wasteful sloths as well. ;)
12) Message boards : Number crunching : Initial Replication (Message 17676)
Posted 1 Aug 2007 by Profile [B^S] Astral Walker
Post:
I am assuming that one day in the not too distant future there will be a steady supply of LHC work or that the runs will become so big (or hosts so few) they won't be gobbled up in a few hours.

While I haven't been with the project that long I don't remember this being true in the past and I wouldn't hold my breath that it will ever be true.
13) Message boards : Number crunching : Question about WU credits (Message 17547)
Posted 25 Jul 2007 by Profile [B^S] Astral Walker
Post:
It appears that the work returned by your computer did not match well enough with what the others had, so the validator marked it invalid.

Thanks. I hope my next CPU gets better grades than this one.
14) Message boards : Number crunching : Question about WU credits (Message 17517)
Posted 23 Jul 2007 by Profile [B^S] Astral Walker
Post:
I've never asked a question about credits before and I rarely look at them, even on projects which provide work routinely. However, a system I attached a few months ago got WUs for the first time and when I checked to see that everything was running smoothly, I noticed that one of my WUs which ran for a few hours was granted 0 credits.

The WU in question is 1690554. When I looked at it in detail, I see that the other users who completed the WU did receive credit. Is there anything I should be concerned about here?



©2021 CERN