Message boards : Number crunching : Validate Windows v. Linux machines?
Message board moderation

To post messages, you must log in.

AuthorMessage
Jim1348

Send message
Joined: 28 Feb 15
Posts: 253
Credit: 200,562,581
RAC: 0
Message 551 - Posted: 21 Sep 2015, 19:10:55 UTC

I see that all my results on a Win7 64-bit machine are being validated against other Windows machines. I don't see any Linux PCs in the group, which seems unlikely considering that over half the GigaFLOPS are being provided by Linux machines.

Do the Windows machines produce different results than the Linux machines, and so have to be validated only against other Windows machines?
ID: 551 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Krzysztof Piszczek - wspieram ...
Project administrator
Project developer
Project tester
Avatar

Send message
Joined: 4 Feb 15
Posts: 847
Credit: 144,180,465
RAC: 0
Message 552 - Posted: 21 Sep 2015, 19:26:23 UTC - in response to Message 551.  
Last modified: 21 Sep 2015, 19:26:43 UTC

The difference is in end line coding in text files between Linux and Windows machines, so same results file for these two systems are different in validator point of view...
Krzysztof 'krzyszp' Piszczek

Member of Radioactive@Home team
My Patreon profile
Universe@Home on YT
ID: 552 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Mumps [MM]

Send message
Joined: 7 Apr 15
Posts: 13
Credit: 136,390,419
RAC: 0
Message 553 - Posted: 22 Sep 2015, 2:46:57 UTC - in response to Message 552.  

And to further clarify the answer, I believe that indicates you have enabled

https://boinc.berkeley.edu/trac/wiki/HomogeneousRedundancy

Probably with the "Coarse grain" setting?

Wiki wrote:
An "HR type" is a host classification. Currently the following HR types are defined:

0
No homogeneous redundancy (all hosts are numerically equivalent)
1
A fine-grained classification with 80 classes (4 OS and 20 CPU types).
2
A coarse-grained classification in which there are 4 classes: Windows, Linux, Mac-PPC and Mac-Intel.

So WU's are only sent to "similar" hosts, so the validation doesn't have to worry about the line ending variations. Is that correct?
ID: 553 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Krzysztof Piszczek - wspieram ...
Project administrator
Project developer
Project tester
Avatar

Send message
Joined: 4 Feb 15
Posts: 847
Credit: 144,180,465
RAC: 0
Message 555 - Posted: 22 Sep 2015, 15:09:24 UTC - in response to Message 553.  

Yes, you are correct :)

We use type 2.
Krzysztof 'krzyszp' Piszczek

Member of Radioactive@Home team
My Patreon profile
Universe@Home on YT
ID: 555 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote

Message boards : Number crunching : Validate Windows v. Linux machines?




Copyright © 2024 Copernicus Astronomical Centre of the Polish Academy of Sciences
Project server and website managed by Krzysztof 'krzyszp' Piszczek