Page 1 of 3 123 LastLast
Results 1 to 30 of 80

Thread: Metacompressor.com benchmark updates

  1. #1
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    782
    Thanks
    63
    Thanked 273 Times in 191 Posts

    Metacompressor.com benchmark updates

    Metacompressor.com was started as big files speed benchmark, to offer more diversity I added today an option to add your own test files. Maximal 3 each user and max 10MB zipped.

    First login and there click "add test file":
    http://www.metacompressor.com/login.aspx

    New users can register here:
    http://www.metacompressor.com/newuser.aspx

    Archivers can be uploaded here:
    http://www.metacompressor.com/submit.aspx

    Results can be watched here:
    http://www.metacompressor.com/uploads.aspx

    Code: LongNightsDebugging

  2. #2
    Programmer osmanturan's Avatar
    Join Date
    May 2008
    Location
    Mersin, Turkiye
    Posts
    651
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Thumbs up

    Thanks sportman! Step by step, you are building a huge benchmark portal

  3. #3
    Tester
    Nania Francesco's Avatar
    Join Date
    May 2008
    Location
    Italy
    Posts
    1,565
    Thanks
    220
    Thanked 146 Times in 83 Posts

    Hi !

    Surely a test very useful and complete! Very Good job!

  4. #4
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    782
    Thanks
    63
    Thanked 273 Times in 191 Posts
    After selecting a test file or uploading a test file it's now possible to select an archiver from a list of archivers instead of uploading for each test an archiver:
    http://www.metacompressor.com/submit.aspx

    Code: LongNightsDebugging

  5. #5
    Programmer osmanturan's Avatar
    Join Date
    May 2008
    Location
    Mersin, Turkiye
    Posts
    651
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Thumbs up

    That's what I need! Thanks!

  6. #6
    Tester
    Nania Francesco's Avatar
    Join Date
    May 2008
    Location
    Italy
    Posts
    1,565
    Thanks
    220
    Thanked 146 Times in 83 Posts

    Hi !

    Good change!

  7. #7
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    782
    Thanks
    63
    Thanked 273 Times in 191 Posts
    After selecting an archiver it's now possible to select command line parameters from a list, manual typing is also still possible:
    http://www.metacompressor.com/submit.aspx

    Code: LongNightsDebugging

  8. #8
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    782
    Thanks
    63
    Thanked 273 Times in 191 Posts
    I raised the upload limit for zipped archivers from 1MB to 2MB and the upload limit for zipped test files from 10MB to 15MB.

  9. #9
    Member
    Join Date
    May 2008
    Location
    Antwerp , country:Belgium , W.Europe
    Posts
    487
    Thanks
    1
    Thanked 3 Times in 3 Posts
    It seems like "metacompressor.com" is broken : after 6 days, my request is still "waiting" and doesn't show up in the list.
    After logging in, I still can see it, but no results yet
    Any clues ?

  10. #10
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    782
    Thanks
    63
    Thanked 273 Times in 191 Posts
    I unfroze the server by taking it away from the Iceland bank deposit safe to a hotter and saver climate.

    No the opposite was true, some archivers where uploaded who started a paq8ox variant in the background that run for ever because paq8ox variants try to make a temp file in root C:\ and that's not allowed under Windows Vista and Windows 2008 Server and fail.

    I stopped this endless paq8ox loop and the queued archiver tests one by one finished. I didn't saw queued tests older then one day, maybe the one Pat wait for is a failed one. To make it more clear, I added a status description to the archiver test results displayed after login.

    I also added a new option that uploaded new archivers can be stored (if enabled) so others can test with it too.
    Last edited by Sportman; 10th October 2008 at 00:38.

  11. #11
    Member
    Join Date
    Jun 2008
    Location
    USA
    Posts
    111
    Thanks
    0
    Thanked 0 Times in 0 Posts
    Quote Originally Posted by Sportman View Post
    No the opposite was true, some archivers where uploaded who started a paq8ox variant in the background that run for ever because paq8ox variants try to make a temp file in root C:\ and that's not allowed under Windows Vista and Windows 2008 Server and fail.
    Vista etc. don't let you run as Admin by default, and for whatever reason, MSVCRT's tmpfile() tries writing to the root dir by default (doh!). And MinGW (at least current stable, last I checked) still uses such, which is why most paq8ox variants fail. However, if you recompile with Cygwin, OpenWatcom, DJGPP, etc. it will work okay. (That's what I did for paq8o8z.) Of course, the DOS ones won't run under x86-64, but that's (mostly?) AMD's fault. (I prefer DJGPP because it has G++ 4.3.2, which is better than 3.4.x.) :-P

  12. #12
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    782
    Thanks
    63
    Thanked 273 Times in 191 Posts
    Thanks, I tested paq8o8zw and it work fine under Windows 2008.

    I noticed a strange problem, some archivers like CMM4 stop using CPU after 5-20 seconds and the captured output text also stop showing new data but the process run forever. When I run it manual at the DOS prompt it work fine, it's if the archiver has a problem with being captured by the windows test control application.
    Last edited by Sportman; 12th October 2008 at 06:57.

  13. #13
    Programmer osmanturan's Avatar
    Join Date
    May 2008
    Location
    Mersin, Turkiye
    Posts
    651
    Thanks
    0
    Thanked 0 Times in 0 Posts
    @Sportman:
    Seems you are not using "<machineKey>" tag in your web.config. If you don't use it, every time you will be faced with this error: "Validation of viewstate MAC failed". If you don't know how can you fix this, I can send you a validation key via mail or PM.

  14. #14
    Programmer toffer's Avatar
    Join Date
    May 2008
    Location
    Erfurt, Germany
    Posts
    587
    Thanks
    0
    Thanked 0 Times in 0 Posts
    @Sportman
    Do you use Vista? I used MinGW to compile CMM4, maybe that could be a problem - btw i never had problems like that?! Does it still output something (into the compressed file)?

  15. #15
    Programmer osmanturan's Avatar
    Join Date
    May 2008
    Location
    Mersin, Turkiye
    Posts
    651
    Thanks
    0
    Thanked 0 Times in 0 Posts
    @toffer:
    As you know, I use Vista but I didn't get any error when using your compressors. Sportman's machine has Windows 2008 Server x64 which is very similar Vista in architectural point of view. I think, the problem is about server permissions and/or his automated script/program.

  16. #16
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    782
    Thanks
    63
    Thanked 273 Times in 191 Posts
    Quote Originally Posted by osmanturan View Post
    @Sportman:
    Seems you are not using "<machineKey>" tag in your web.config. If you don't use it, every time you will be faced with this error: "Validation of viewstate MAC failed".
    I generated a machineKey and added it to web.config, let me know if it help.

  17. #17
    Programmer osmanturan's Avatar
    Join Date
    May 2008
    Location
    Mersin, Turkiye
    Posts
    651
    Thanks
    0
    Thanked 0 Times in 0 Posts
    Quote Originally Posted by Sportman View Post
    I generated a machineKey and added it to web.config, let me know if it help.
    It should help. I faced this error in my web sites 1-2 years ago too. machineKey solved all of related problems. Let me give 20-30 minutes for ensuring whether the problem has been fixed or not. As you know, ASP.NET flushes the session data after this default time out value unless you didn't override a new time out value in web.config.

  18. #18
    Programmer osmanturan's Avatar
    Join Date
    May 2008
    Location
    Mersin, Turkiye
    Posts
    651
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Thumbs up

    I have tested again. Seems the problem is fixed.

  19. #19
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    782
    Thanks
    63
    Thanked 273 Times in 191 Posts
    Quote Originally Posted by toffer View Post
    @Sportman
    Do you use Vista? Does it still output something (into the compressed file)?
    Windows 2008 64bit what is already mentioned is almost similar as a cleaned Windows Vista.

    I did some testing to see when it goes wrong:

    file14 - 5,629,573 bytes:
    628,073 bytes 3.134 sec. - cmm 54 @infile @outfile cmm d @infile @outfile

    file15 - 160,340 bytes:
    65,608 bytes 0.256 sec. - cmm 54 @infile @outfile cmm d @infile @outfile

    file16 - 7,882,752 bytes:
    926,341 bytes 4.49 sec. - cmm 54 @infile @outfile cmm d @infile @outfile

    file17 - 1,145,452 bytes:
    341,316 bytes 1.024 sec. - cmm 54 @infile @outfile cmm d @infile @outfile

    file18 - 8,009,113 bytes:
    522,342 bytes 3.929 sec. - cmm 54 @infile @outfile cmm d @infile @outfile

    file20 - 7,881,296 bytes:
    4,202,005 bytes 6.753 sec. - cmm 54 @infile @outfile cmm d @infile @outfile

    file23 - 3,352,576 bytes:
    345,557 bytes 1.61 sec. - cmm 54 @infile @outfile cmm d @infile @outfile

    file24 - 7,424,946 bytes:
    7,220,160 bytes 7.785 sec. - cmm 54 @infile @outfile cmm d @infile @outfile

    This tests run fine, when the test file size goes up the problem started:

    file13 - 10,485,760 bytes:
    After 10.25 seconds CPU time CPU usage go to zero
    Output is 9,732,096 bytes and not increasing any more
    When I end cmm.exe I see last four rows captured command line output where:
    Encoding: 9122/ 9088kB (8.03 bpc). 934 kB/s
    Encoding: 9250/ 9216kB (8.03 bpc). 934 kB/s
    Encoding: 9379/ 9344kB (8.03 bpc). 935 kB/s
    Encoding: 9507/

    file19 - 12,305,247 bytes:
    After 5.27 seconds CPU time CPU usage go to zero
    Output is 573,440 bytes and not increasing any more
    When I end cmm.exe I see last four rows captured command line output where:
    Encoding: 541/ 9088kB (0.48 bpc). 1831 kB/s
    Encoding: 548/ 9216kB (0.48 bpc). 1834 kB/s
    Encoding: 555/ 9344kB (0.48 bpc). 1831 kB/s
    Encoding: 563/

    file27 - 14,357,084 bytes:
    After 9.67 seconds CPU time CPU usage go to zero
    Output is 7,270,400 bytes and not increasing any more
    When I end cmm.exe I see last four rows captured command line output where:
    Encoding: 6819/ 9088kB (6.00 bpc). 992 kB/s
    Encoding: 6914/ 9216kB (6.00 bpc). 991 kB/s
    Encoding: 7008/ 9344kB (6.00 bpc). 991 kB/s
    Encoding: 7103/

    At all three failed tests the last four command line output rows show 9088kB, 9216kB, 9344kB and freeze just before printing I guess 9344kB + 128kB = 9472kB.
    Last edited by Sportman; 15th October 2008 at 16:47.

  20. #20
    Programmer toffer's Avatar
    Join Date
    May 2008
    Location
    Erfurt, Germany
    Posts
    587
    Thanks
    0
    Thanked 0 Times in 0 Posts
    But all of that is still related to your automated testing? Not "manual" running at the command prompt. I might recompile it using a different compiler. ATM i cannot imagine any bugs which cause stuff like that. Let me think a bit about that...

    BTW what other archivers suffer from such a behavior? Maybe it's a MinGW related problem?

  21. #21
    Programmer Bulat Ziganshin's Avatar
    Join Date
    Mar 2007
    Location
    Uzbekistan
    Posts
    4,497
    Thanks
    735
    Thanked 660 Times in 354 Posts
    my programs are compiled using mingw. freearc, 4x4, tornado. but i don't use tmpfile()

  22. #22
    Programmer toffer's Avatar
    Join Date
    May 2008
    Location
    Erfurt, Germany
    Posts
    587
    Thanks
    0
    Thanked 0 Times in 0 Posts
    I don't use it, too. Does M1 work for you? If yes, that's one more reason to replace CMM4's I/O system (it should become faster, as a bonus).

    @Bulat
    Which MinGW version did you use?

  23. #23
    Programmer Bulat Ziganshin's Avatar
    Join Date
    Mar 2007
    Location
    Uzbekistan
    Posts
    4,497
    Thanks
    735
    Thanked 660 Times in 354 Posts
    some gcc 3.4.x

  24. #24
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    782
    Thanks
    63
    Thanked 273 Times in 191 Posts
    Quote Originally Posted by toffer View Post
    But all of that is still related to your automated testing? Not "manual" running at the command prompt. I might recompile it using a different compiler. ATM i cannot imagine any bugs which cause stuff like that. Let me think a bit about that...

    BTW what other archivers suffer from such a behavior? Maybe it's a MinGW related problem?
    Yes they have no problem at the command prompt. The automatic test application start every archiver in a new separate thread and this thread poll every second to capture the redirect standard and error output till the archiver is finished.

    So far I saw that archivers CMM, CCM and CCMX has this issue. CCM and CCMX have no problem with file13, file19 and file27 but from file11 or bigger they freeze at similar way.

  25. #25
    Programmer toffer's Avatar
    Join Date
    May 2008
    Location
    Erfurt, Germany
    Posts
    587
    Thanks
    0
    Thanked 0 Times in 0 Posts
    I used GCC 4.2.1 (which produced notable faster code), and what about M1?

  26. #26
    Programmer
    Join Date
    May 2008
    Location
    denmark
    Posts
    94
    Thanks
    0
    Thanked 2 Times in 2 Posts
    Sportman,

    Can't you try to fix the bug for qpress 0.22? I'm eager to see if the new disk I/O (changed after 0.14) works as expected

  27. #27
    Member
    Join Date
    Aug 2008
    Location
    Planet Earth
    Posts
    782
    Thanks
    63
    Thanked 273 Times in 191 Posts
    Quote Originally Posted by toffer View Post
    what about M1?
    I don't know if this question is for me, but M1 had no problem at any test file.

  28. #28
    Member m^2's Avatar
    Join Date
    Sep 2008
    Location
    Ślůnsk, PL
    Posts
    1,611
    Thanks
    30
    Thanked 65 Times in 47 Posts
    Hey Sportman, I have some comments.

    I love the idea, it's so awesome to be able to make own tests on a different machine, but some things could be improved.
    1. It may be just a long queue, but possibly something hung again, I have some tests pending for a day and night. Could you look into it? And if such long queues happen, it would be nice if you showed how long (preferably in gigabytes) it is.
    2. You have too many results to browse comfortably and things will be worse. Could you implement some filtering, i.e. to see only the latest archiver version or to remove failed modes?
    I think that option to export to csv would be great too (probably even better).
    3. I'd welcome a column separator.
    4. How do you measure times? Some results look strange, i.e. file2, FreeArc -mdelta+rep:800mb+rzm has 6s decompression CPU time. rzm - 71.5s. rep can help, possibly delta too, but....

  29. #29
    Member m^2's Avatar
    Join Date
    Sep 2008
    Location
    Ślůnsk, PL
    Posts
    1,611
    Thanks
    30
    Thanked 65 Times in 47 Posts
    It's working now, but I have another request: Could you increase maximum command line length? I'm short of ~10 chars.

  30. #30
    Member
    Join Date
    Dec 2008
    Location
    Poland, Warsaw
    Posts
    940
    Thanks
    558
    Thanked 379 Times in 283 Posts

    Question Problem with paq8k submission

    H!,
    I've encountered some problems with adding paq8k version to metacompressor benchmarks.
    I've tried to add paq8k compressor to "uiq2" benchmark, however after finishing submission steps and uploading archiver, I've got status message "Compression finished: 'uiq2' is compressed to 0 bytes in 0 seconds". I've checked out details and found that compressed file has 29 bytes - it's adequate value to paq8k header.
    It looks like paq8k compressor run, but then, due to some reasons, it stopped and reports error.

    What could be wrong? Command line looks proper. When I could find information about error reported by compressor?

    Could you suggest me what should I do to run this benchmark Properly?

    Regards, Darek

Page 1 of 3 123 LastLast

Similar Threads

  1. zpaq updates
    By Matt Mahoney in forum Data Compression
    Replies: 2527
    Last Post: 4th May 2019, 12:33
  2. Metacompressor.com first contest
    By Sportman in forum Data Compression
    Replies: 17
    Last Post: 15th October 2008, 00:50

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •