Einstein problems

Forum rules
sneakysaurus

#1 Einstein problems

Post by sneakysaurus »

Can anyone help me with a major Einstein problem I have?

Basically every task I download immediately reports computation error!

I completely reinstalled my pc this week getting rid of three crappy hard disks and replacing with a pair of 500Gb SATA drives. Now running Ubuntu 8.04 64Bit and Vista 64 bit.

Installed Ubuntu without issues as before but am now getting this problem with Einstein. WCG and Rosetta work fine so I am very confused. Tried a manual install with the .sh script and also installed via Synaptics manager but both give me the same result. Any ideas? Or any ideas where to investigate or what log files to look at?

Not too fussed just now since will just stick with WCG and Rosetta just but really want to sort this since I was racking up a nice amount of credit running Einstein on Linux.

I am not over clocking. I saw this problem once a few weeks back when pushing my q6600 to 3.15Ghz but back at stock 2.4 now so really confused. I thought it may be a permission problem but surely all projects would suffer same result?

An example task result is this:

http://einstein.phys.uwm.edu/result.php ... d=97491120

Any help is greatly appreciated.
Nightlord

#2

Post by Nightlord »

32 bit libraries?

If I recall Ubuntu64 doesn't have them pre-installed and a number of apps require them on 64 bit distro's. Try use Synaptic Package Manager to install ia32libs (or similar name).

Trouble is you may have used you quota of 16 WU for today, so if necessary, after installing the 32bit libraries, disconnect and re-connect to force a new installation and download a new batch of WU.

Good luck!
Matelot

#3

Post by Matelot »

I've tried to search for the 32libs but synaptic Package Manager cant see any, probably my fualt for being a linux greenhorn
Nightlord

#4

Post by Nightlord »

meely-mouthed get out clause: the following was taken from an Ubuntu 7.04 Live session, it may be different on the installed version.

In synaptic package manager, search for "ia32-libs". It will probably find several dependencies that need installed alongside.

Still not sure if that was sneaky's problem, but it's worth a punt.
sneakysaurus

#5

Post by sneakysaurus »

It worked!

Cheers Nightlord for the good advice. I would have really struggled to get this going myself.

I just searched for ia32 in synaptic package manager. I installed ia32-libs and dependencies and it all worked.

I didn't to do this the last time I installed Ubuntu but maybe I installed something else that required these package.
Anyway one to remember for future issues.

Thanks again.
Matelot

#6

Post by Matelot »

Ditto :)

Cheers Nightlord, got the libs, and I'm up and running :D :D
Matelot

#7

Post by Matelot »

Just noticed that I have a lot of "Validate Errors", I take it I won't get any credit for these? if not, I've lost a load of time and credits because of the einstein server :crybaby:

98236617 39796802 26 May 2008 17:10:14 UTC 30 May 2008 8:05:39 UTC Over Validate error Done 28,739.08 --- ---
98236607 39796797 26 May 2008 17:10:12 UTC 30 May 2008 8:05:39 UTC Over Validate error Done 28,671.84 --- ---
98236573 39796780 26 May 2008 17:10:03 UTC 30 May 2008 8:05:39 UTC Over Validate error Done 29,904.51 --- ---
98236561 39796774 26 May 2008 17:10:01 UTC 29 May 2008 6:33:12 UTC Over Validate error Done 30,152.07 --- ---
98236492 39796742 26 May 2008 17:08:51 UTC 29 May 2008 6:33:12 UTC Over Validate error Done 28,592.82 --- ---
98236484 39796738 26 May 2008 17:08:47 UTC 29 May 2008 6:33:12 UTC Over Validate error Done 28,176.52 --- ---
98236482 39796737 26 May 2008 17:08:45 UTC 29 May 2008 6:33:12 UTC Over Validate error Done 27,985.32 --- ---
98236480 39796736 26 May 2008 17:08:43 UTC 29 May 2008 6:33:12 UTC Over Validate error Done 27,789.37 --- ---
98236476 39796734 26 May 2008 17:08:41 UTC 29 May 2008 6:33:12 UTC Over Validate error Done 27,400.69 --- ---
98236472 39796732 26 May 2008 17:08:39 UTC 28 May 2008 6:30:44 UTC Over Validate error Done 27,737.45 --- ---
98236466 39796729 26 May 2008 17:08:37 UTC 28 May 2008 6:30:44 UTC Over Validate error Done 27,227.00 --- ---
98236464 39796728 26 May 2008 17:10:08 UTC 30 May 2008 8:05:39 UTC Over Validate error Done 24,716.80 --- ---
98236456 39796724 26 May 2008 17:10:08 UTC 30 May 2008 8:05:39 UTC Over Validate error Done 25,174.96 --- ---
98236454 39796723 26 May 2008 17:09:59 UTC 29 May 2008 6:33:12 UTC Over Validate error Done 30,522.98 --- ---
98236446 39796719 26 May 2008 17:10:06 UTC 30 May 2008 8:05:39 UTC Over Validate error Done 28,174.94 --- ---
98236435 39796714 26 May 2008 17:08:28 UTC 28 May 2008 6:30:44 UTC Over Validate error Done 30,061.76 --- ---
98235868 39796442 26 May 2008 17:08:33 UTC 28 May 2008 6:30:44 UTC Over Validate error Done 25,611.52 --- ---
98233641 39210791 26 May 2008 17:08:26 UTC 28 May 2008 6:30:44 UTC Over Validate error Done 26,025.84 --- ---
98233531 39795326 26 May 2008 17:08:30 UTC 28 May 2008 6:30:44 UTC Over Validate error Done 27,329.08 --- ---
utopia-i
Boinc Warrant Officer Class 2
Boinc Warrant Officer Class 2
Posts: 338
Joined: Fri Sep 14, 2007 5:00 pm
Location: Edinburgh

#8

Post by utopia-i »

the latest news .
News items

Jul 22, 2008
We are reaching the end of the current analysis run and expect the last workunits of S5R3 to be generated around Aug 1. Participants will probably experience an increase of the downloads necessary for the remaining S5R3 workunits to be processed. Users with low bandwidth will probably want to suspend the project for some days. Users of Beta- and Power-Apps should be aware that they will not get tasks of the upcomig S5R4 run with these S5R3 Apps.
I don't do beta but what are the "Power-Apps"?
... is that apple pie I smell

Image
utopia-i
Boinc Warrant Officer Class 2
Boinc Warrant Officer Class 2
Posts: 338
Joined: Fri Sep 14, 2007 5:00 pm
Location: Edinburgh

#9

Post by utopia-i »

just discovered this post on Einstein@home - they have server/upgrade issues just know ..
http://einstein.phys.uwm.edu/forum_thread.php?id=6758
Another \"incentive\" to stay for the cleanup will be the fact that old work will very likely pay a lot more than the new :-). The new will have the \"downgraded\" pay scales that Bernd was talking about some time ago :-). That sounds like a good plan to me - do the \"cr*p\" work and be paid accordingly :-).
so both Einstein and Cosmology are going thru a global credit crunch - perhaps some BOINC project with raise the WU credit following of from Scottish Gas :P
... is that apple pie I smell

Image
Nightlord

#10

Post by Nightlord »

The Einstein server upgrade work should be completed today (according to their plan last week). As I type this though, I'm still having upload problems. It'll work itself out, then think of the credit 8)

......unlike Cosmo that granted me zero credit for many hours crunching just because my wingman happened to be assigned the new app while I was assigned the old. Apparently new and old apps are not compatible. I think it was about a dozen or so WU's across a range of boxes, each one lasting several hours each before I pulled the plug. Will check back in a few weeks. :evil:
Nightlord

#11

Post by Nightlord »

The server is up and I finally managed to report all the WU's, but there was a hiccup during the migration process. For several hours there was a problem with uploads that trashed WU's. SO far I've lost 16 :(

OK, so this has been a very stable and responsive project. And the project lead scientist personally comes to the forum to apologise (see here). So, I'm prepared to live and let live. At least there is feedback, rather than a deafening silence such as you get from other projects that trash your WU's :roll:
utopia-i
Boinc Warrant Officer Class 2
Boinc Warrant Officer Class 2
Posts: 338
Joined: Fri Sep 14, 2007 5:00 pm
Location: Edinburgh

#12

Post by utopia-i »

and see here for the new credit [issues?]

http://einstein.phys.uwm.edu/forum_thread.php?id=6806

http://einstein.phys.uwm.edu/forum_thread.php?id=6802

the basics, OK on 1 result only! ...

Linux

old = 4 hours 237 points given = 59 points per hour

new = 6.3 hours 192.24 given = 30.51 points per hour :roll: :?: :!:

50% less :twisted: :evil:
... is that apple pie I smell

Image
User avatar
Megacruncher
G.L.S.B.
G.L.S.B.
Posts: 4699
Joined: Mon May 29, 2006 11:33 pm
Location: Edinburgh, Scotland
Contact:

#13

Post by Megacruncher »

Mmm, there's a pattern emerging here what with an even more brutal "correction" on cosmology and rieselsieve going offline. The good payers are being reined in/ self-regulating and we are in serious danger of WCG being averagely rewarding. Of course it's how Boinc should be but a bit boring. :(
Willie the Megacruncher
Image
Nightlord

#14

Post by Nightlord »

I've not posted about this over at Einstein for fear of being black listed or otherwise condemned with sideways glances.

But here I think I can speak a bit easier.

There are two things that really annoy me about this whole credit adjustment......

Firstly, there appears to be a small group of people going round each high paying project telling them to decrease their credit rates, but I see no evidence of the reverse - pressure to increase badly paying projects.

Secondly, projects like Einstein have just converted over to using "optomised" science apps as standard build, yet they are compared in the credit rates with stock apps at Seti. It is clear to me that if I wanted to reach a high project RAC, I will have to go back to Seti and run optomised apps.

There are a few alternatives, I could go back to ABC and run true 64bit apps that crunch twice as fast as 32bit, or I could increase the commitment on GPUGrid (PS3Grid). Meanwhile, I'm pushing on to 1M on this project while looking at QMC and Milkyway.
User avatar
Megacruncher
G.L.S.B.
G.L.S.B.
Posts: 4699
Joined: Mon May 29, 2006 11:33 pm
Location: Edinburgh, Scotland
Contact:

#15

Post by Megacruncher »

The puritan tendancy is a bit irksome. As you say it's all about reducing credit rather than equalising it. :(
On the whole though this Boinc credit crunch is a good, if rather boring, thing. If each project was always trying to be more generous than every other one, and got off with it, then we would have run away inflation, Zimbabwe style, and soon our hard gained 92 million amassed over 4 years would be acheivable by a low end laptop running Cosmo@home for an afternoon. :shock:
Anyhoo I'm going to use my belated discovery of the optimised Seti client to get my Seti million, boost my RAC (and that of TSBT) and delay the inevitable serial demotions which are my future prospects!
Willie the Megacruncher
Image
User avatar
Buster Gunn
Boinc Second Lieutenant
Boinc Second Lieutenant
Posts: 513
Joined: Sun Aug 13, 2006 7:20 am
Location: Wilmington Delaware USA

#16

Post by Buster Gunn »

I just installed the optomized client on my machines. What a buzz. It even gets rid of those unwanted Astropulse w/u's.
Buster is my dog. Sadly, Buster is gone.
Buster is the Malamute, Obi is the Golden.

Image
Obi is the dog of the house now.
utopia-i
Boinc Warrant Officer Class 2
Boinc Warrant Officer Class 2
Posts: 338
Joined: Fri Sep 14, 2007 5:00 pm
Location: Edinburgh

#17

Post by utopia-i »

Nightlord wrote:I've not posted about this over at Einstein for fear of being black listed or otherwise condemned with sideways glances.
Now you tell me! :P Lets just say there are more friendly BOINC projects out there I'll be focusing on ...

but progress .....

Credit adjustment message from project Developer
http://einstein.phys.uwm.edu/forum_thre ... 6815#86438

and I can confirm!

New WUs
1st run 21,190.32 65.83 192.24
2nd run 20,566.95 65.83 220.54 <==== better!

prevous was 14,431.18 237.42 237.42
... is that apple pie I smell

Image
sneakysaurus

#18

Post by sneakysaurus »

My first results on new credit rating was:

http://einstein.phys.uwm.edu/workunit.php?wuid=42231641

19,528.28secs for 64.16 credits , now that is total sheeetye!!


I used to get:

http://einstein.phys.uwm.edu/workunit.php?wuid=42040946

14,128.69secs for 237.42 credits.

Quite a BIG difference. I have binned Einstein for now but mught try again in a few weeks.
utopia-i
Boinc Warrant Officer Class 2
Boinc Warrant Officer Class 2
Posts: 338
Joined: Fri Sep 14, 2007 5:00 pm
Location: Edinburgh

#19

Post by utopia-i »

sneakysaurus wrote:My first results on new credit rating was:

http://einstein.phys.uwm.edu/workunit.php?wuid=42231641

19,528.28secs for 64.16 credits , now that is total sheeetye!!


I used to get:

http://einstein.phys.uwm.edu/workunit.php?wuid=42040946

14,128.69secs for 237.42 credits.

Quite a BIG difference. I have binned Einstein for now but mught try again in a few weeks.
Thats your claimed credit - your granted credit will be 192.nn or hopefully 220.nn once the quorum is completed.
... is that apple pie I smell

Image
User avatar
McShane of TSBT
Boinc Private
Boinc Private
Posts: 12
Joined: Thu Apr 27, 2023 12:12 am

#20 Re: Einstein problems

Post by McShane of TSBT »

I don't seem to be able to join the Team in this project. I get an answer that I do not have permission access this resource. I made a note in their forum and all I got back so far is that the Team might be closed to new members. This does not seem right. Aby clues?
User avatar
scole of TSBT
Boinc Major General
Boinc Major General
Posts: 5980
Joined: Mon Feb 03, 2014 2:38 pm
Location: Goldsboro, (Eastern) North Carolina, USA

#21 Re: Einstein problems

Post by scole of TSBT »

@Megacruncher can you check this out? You are the founder on Einstein. Check to make sure we're accepting new members
Image
User avatar
Granowski
Marathon Man, Morale Officer, Prime Mover
Marathon Man, Morale Officer, Prime Mover
Posts: 747
Joined: Wed Aug 21, 2019 3:35 pm
Location: Wrocław, Poland
Contact:

#22 Re: Einstein problems

Post by Granowski »

with Einstein I am over, with 100 mln points ... regards, Gregg Roman Granowski
Image
Post Reply Previous topicNext topic

Return to “Einstein”