[opensuse-factory] Disk transaction performance - Phoronix benchmarking
Hi all. Phoronix has run distribution comparison [1]. I´m just wondering why disk transaction performance has Ubuntu 10.04 faster than openSUSE 11.3 RC1, when both distros uses an ext4 file system. Does openSUSE have enabled any barriers again, as was in openSUSE 11.1? http://www.phoronix.com/scan.php?page=article&item=linux_2010_fiveway&num=1 Thanks. -- S pozdravom / Best regards, Rasto -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
Le lundi 28 juin 2010, à 10:29 -0400, Cristian Rodríguez a écrit :
So let's have a blog post or news story on news.opensuse.org to explain our position! The right thing is not to ignore the benchmarks, but to educate people so we don't look bad with no reason. Vincent -- Les gens heureux ne sont pas pressés. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 [sent later] On 2010-06-28 16:40, Vincent Untz wrote:
I don't take them as the ultimate truth, but neither as BS.
Meanwhile, why doesn't any of you comment on it? Stephan has hinted on functionality and data integrity; perhaps we are using different defaults than others, but knowing what those could be would be nice. Perhaps it is something as simple as the comparison being run against a beta with debug capabilities enabled :-? - -- Cheers / Saludos, Carlos E. R. (from 11.2 x86_64 "Emerald" GM (Minas Tirith)) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.12 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org/ iF4EAREIAAYFAkwpjHQACgkQja8UbcUWM1wkewD+N6zqZhphJdj1xthWxXsC2yH6 5t11W+VjNyfkJVw77EsA/0SBiy5u+BCEzGlO+45DiV0E4mfyXfKByq3BX7AQKtHh =A25C -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
On Tuesday 29 June 2010 06:02:28 Carlos E. R. wrote:
No matter whether you might think the benchmarks are BS or not - I cannot answer that technically - this is one of our problems - perception... So, we need to either publicly demonstrate with repeatable test cases that the benchmarks are indeed problematic or find why we are not getting the same results if indeed it is a valid benchmark. I met Michael at LinuxTag and he is a sensible guy who has no bias against openSUSE. I explained in detail why I openSUSE is my preferred distro with some real world examples of success. He was floored when I mentioned the uptimes of some my client's servers on SL*/openSUSE. - years, not months... We have a great distro, but often we have difficulty to explain to the larger public our technical excellence ;) Just my 0.02 $currency Peter -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
On 06/28/2010 08:29 AM, Cristian Rodríguez wrote:
Phoronix benchmarks are BS, I hope you are not taking that thing seriously do you ?
Would you care to post what you find BS about them? Most of the benchmarks are done by using real life applications doing their tasks unlike most other BS benchmarks that live off of synthetic benchmarks which have no relevance in real life applications. Of course if you know of a flaw please feel free to contribute. Patches are welcome. Dean -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
Dňa 28.06.2010 11:12, Rastislav Krupanský wrote / napísal(a):
And again, Ubuntu 10.04 has faster disk transaction performance with older kernel than final 11.3. Can anyone explain why? http://www.phoronix.com/scan.php?page=article&item=opensuse_113_benchmarks&n... Thanks. -- S pozdravom / Best regards, Rasto -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
Le lundi 28 juin 2010, à 10:29 -0400, Cristian Rodríguez a écrit :
So let's have a blog post or news story on news.opensuse.org to explain our position! The right thing is not to ignore the benchmarks, but to educate people so we don't look bad with no reason. Vincent -- Les gens heureux ne sont pas pressés. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 [sent later] On 2010-06-28 16:40, Vincent Untz wrote:
I don't take them as the ultimate truth, but neither as BS.
Meanwhile, why doesn't any of you comment on it? Stephan has hinted on functionality and data integrity; perhaps we are using different defaults than others, but knowing what those could be would be nice. Perhaps it is something as simple as the comparison being run against a beta with debug capabilities enabled :-? - -- Cheers / Saludos, Carlos E. R. (from 11.2 x86_64 "Emerald" GM (Minas Tirith)) -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.12 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org/ iF4EAREIAAYFAkwpjHQACgkQja8UbcUWM1wkewD+N6zqZhphJdj1xthWxXsC2yH6 5t11W+VjNyfkJVw77EsA/0SBiy5u+BCEzGlO+45DiV0E4mfyXfKByq3BX7AQKtHh =A25C -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
On Tuesday 29 June 2010 06:02:28 Carlos E. R. wrote:
No matter whether you might think the benchmarks are BS or not - I cannot answer that technically - this is one of our problems - perception... So, we need to either publicly demonstrate with repeatable test cases that the benchmarks are indeed problematic or find why we are not getting the same results if indeed it is a valid benchmark. I met Michael at LinuxTag and he is a sensible guy who has no bias against openSUSE. I explained in detail why I openSUSE is my preferred distro with some real world examples of success. He was floored when I mentioned the uptimes of some my client's servers on SL*/openSUSE. - years, not months... We have a great distro, but often we have difficulty to explain to the larger public our technical excellence ;) Just my 0.02 $currency Peter -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org
participants (8)
-
Carlos E. R.
-
Cristian Rodríguez
-
Dean Hilkewich
-
Marcus Meissner
-
P Linnell
-
Rastislav Krupanský
-
Stephan Kulow
-
Vincent Untz