Message boards : Number crunching : A lot of mismatching results!
Message board moderation

To post messages, you must log in.

AuthorMessage
metalius
Avatar

Send message
Joined: 3 Oct 06
Posts: 101
Credit: 8,985,206
RAC: 43
Message 22930 - Posted: 7 Sep 2011, 16:44:02 UTC

Hello!

Looking to current validation of my results, the situation for LHC@home is never seen before:
30 results - validated;
8 - validation inconclusive.
So, success rate is less than 4/5!

Regards!
ID: 22930 · Report as offensive     Reply Quote
[AF>FAH-Addict.net]toTOW

Send message
Joined: 9 Oct 10
Posts: 77
Credit: 3,671,357
RAC: 0
Message 22935 - Posted: 7 Sep 2011, 19:51:56 UTC

I'm at 50%, and the validated ones were the smallest that lasted only a few seconds :(
ID: 22935 · Report as offensive     Reply Quote
Matthias Lehmkuhl

Send message
Joined: 15 Jul 05
Posts: 23
Credit: 2,052,895
RAC: 1,302
Message 22936 - Posted: 7 Sep 2011, 19:55:52 UTC
Last modified: 7 Sep 2011, 19:56:49 UTC

got also one Completed, validation inconclusive
http://lhcathomeclassic.cern.ch/sixtrack/workunit.php?wuid=12883
Linux EM64T with wingman Win7 64bit
Matthias

ID: 22936 · Report as offensive     Reply Quote
Profile Nawiedzony
Avatar

Send message
Joined: 4 Jul 07
Posts: 17
Credit: 35,310
RAC: 0
Message 22939 - Posted: 7 Sep 2011, 20:15:33 UTC - in response to Message 22936.  
Last modified: 7 Sep 2011, 20:17:58 UTC

12xWU finished (on winXP32bit app530.08) 8xWU got status:
Completed, validation inconclusive

it doesn't look too good... http://lhcathomeclassic.cern.ch/sixtrack/results.php?userid=79108
ID: 22939 · Report as offensive     Reply Quote
metalius
Avatar

Send message
Joined: 3 Oct 06
Posts: 101
Credit: 8,985,206
RAC: 43
Message 22941 - Posted: 7 Sep 2011, 20:19:07 UTC

Update:
34 - validated;
28 - no consensus.
Success rate is approaching to 1/2.
My prognosis - in the next few hours validator will stop at all.
Happy crunching! :-D

ID: 22941 · Report as offensive     Reply Quote
Profile Nawiedzony
Avatar

Send message
Joined: 4 Jul 07
Posts: 17
Credit: 35,310
RAC: 0
Message 22944 - Posted: 7 Sep 2011, 20:54:17 UTC - in response to Message 22941.  
Last modified: 7 Sep 2011, 21:47:11 UTC

I think I noticed something important:
All my WU that got the status valid, are calculated on another PC with the same operating system as mine (winXP)
All my WU that got status inconclusive, are calculated on another PC with a different operating system than mine.

Is it important information?
ID: 22944 · Report as offensive     Reply Quote
Alyx

Send message
Joined: 22 Oct 10
Posts: 1
Credit: 173,295
RAC: 0
Message 22945 - Posted: 7 Sep 2011, 21:20:24 UTC

I've only got 22 validated successfully with about 180 still waiting validation, most of these got the validated inconclusive, some are actually waiting. Bummer is that the machines are hitting max units per day, because so many of them fail.

@Nawiedzony - I've had failures and good ones all on the same machine.
ID: 22945 · Report as offensive     Reply Quote
Profile Igor Zacharov
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 16 May 11
Posts: 79
Credit: 111,419
RAC: 0
Message 22947 - Posted: 7 Sep 2011, 22:19:18 UTC - in response to Message 22944.  

The validation of sixtrack fails for reason not related to numerical comparison.

We are still debugging this. Sometimes the validator fails to open the sixtrack output files, while the files are in the upload directory.
My guess is that it runs out of memory, or it may be related to the virtual memory setup in which we run boinc.

We have prepared a new fresh machine to transfer the sixtrack operations.

In the mean time, I will trim the daily quota to 2/day and max wus to send to 1.
If our server simply cannot chaw all the results that come in, this will give it some relief and # valid results will increase.

skype id: igor-zacharov
ID: 22947 · Report as offensive     Reply Quote
Profile trigggl
Avatar

Send message
Joined: 17 Feb 09
Posts: 22
Credit: 311,184
RAC: 0
Message 22949 - Posted: 8 Sep 2011, 1:56:27 UTC
Last modified: 8 Sep 2011, 1:56:35 UTC

Being the Linux minority is definitely not conducive here. If there were a way to pair Linux only to Linux...
ID: 22949 · Report as offensive     Reply Quote
metalius
Avatar

Send message
Joined: 3 Oct 06
Posts: 101
Credit: 8,985,206
RAC: 43
Message 22951 - Posted: 8 Sep 2011, 7:47:31 UTC
Last modified: 8 Sep 2011, 8:27:47 UTC

Update.
My hosts started to produce Invalid results.
Situation looks like:
1. validator is marking as Invalid all 3rd results after situation "Validation inconclusive";
2. WU is closed with canonical result (_0 or _1) after that.
Please, fix it (if I am right).
ID: 22951 · Report as offensive     Reply Quote
tullio

Send message
Joined: 19 Feb 08
Posts: 708
Credit: 4,336,250
RAC: 0
Message 22954 - Posted: 8 Sep 2011, 10:13:20 UTC

Four wingmen, 2 Wintel valid results, 2 AMD, one Windows7 one Linux (mine), invalid.
Tullio
ID: 22954 · Report as offensive     Reply Quote
mfbabb2

Send message
Joined: 10 Oct 08
Posts: 19
Credit: 7,191
RAC: 0
Message 22961 - Posted: 8 Sep 2011, 16:11:04 UTC

Since most (if not all) of the recent batch of Invalids were due to the Validator itself, will we eventually get credit for the work done; or is it "water under the bridge"?
ID: 22961 · Report as offensive     Reply Quote
Matthias Lehmkuhl

Send message
Joined: 15 Jul 05
Posts: 23
Credit: 2,052,895
RAC: 1,302
Message 22974 - Posted: 9 Sep 2011, 6:43:24 UTC

valid to same computer, my is the invalid also running Win7 x64.
http://lhcathomeclassic.cern.ch/sixtrack/workunit.php?wuid=5730
both valid results where calculated on computer 9915117 Win7 x64.


Matthias

ID: 22974 · Report as offensive     Reply Quote

Message boards : Number crunching : A lot of mismatching results!


©2024 CERN