Skip navigation
electroteque1
Currently Being Moderated

Getting 503 errors for f4f httpstreaming files

Feb 18, 2011 7:00 PM

I am testing out the f4f apache module, two issues have come up but no fix. I have tested packaging up the demo files found with FMS. Every file ive tested obtains a 503 error when it tries to load

 

something like this

 

sample1_700kbpsSeg1-Frag1

 

No proper debug error in the apache logs at all which isn't helpful.

 

the config is setup right because i know its being activated

 

<Location /videos/vod>

    HttpStreamingEnabled true

    HttpStreamingContentPath "/var/www/httpstreaming.electrocode.net/web/videos/vod/f4f"

    Options -Indexes FollowSymLinks

</Location>

I inspected the files and this is what it shows when i run this /opt/adobe/f4fpackager/f4fpackager --input-file=sample1_700kbpsSeg1.f4f --inspect-fragments
It's very wierd, no debug messages at all !
Prepackaged f4f files would be nice to know that the module is actually working correctly.
Let me know.

 
Replies
  • Currently Being Moderated
    Feb 22, 2011 9:12 PM   in reply to electroteque1

    Hi,

     

    The output on inspect-fragment looks correct and it seems file has been packaged correctly. It seems like your F4F module is not installed correctly. Can you verify your logs and see if you get an idea.

     

    Thanks,

    Abhishek

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 23, 2011 4:07 AM   in reply to electroteque1

    Hi,

     

    I do not think the inspect-fragments dump is entirely correct. I have fragmented the same file (the demo file from FMS) with default fragmentation values and was able to playback. I am copying here the first few lines from the inspect fragments dump:

     

    fragment: 1, 74912

    afra:

    long ids: true

    long offsets: true

    entries:

    time = 0, offset = 16511

    time = 3003, offset = 275611

    global entries:

    bootstrap:

    version: 1

    is update: false

    is live: false

    movie id:

    timescale: 1000

    current media time: 6006

    smpte timecode offset: 0

    servers:

    qualities:

    segments:

    segment run table:

    is update: false

    qualities:

    entries:

    segment = 1, fragments = 1

    fragments:

    fragment run table:

    is update: false

    qualities:

    entries:

    fragment = 1, timestamp = 0, duration = 6000

    fragment: 2, 600951

    afra:

    long ids: true

    long offsets: true

    entries:

    time = 6006, offset = 8367

    global entries:

    bootstrap:

    version: 2

    is update: false

    is live: false

    movie id:

    timescale: 1000

    current media time: 9009

    smpte timecode offset: 0

    servers:

    qualities:

    segments:

    segment run table:

    is update: false

    qualities:

    entries:

    segment = 1, fragments = 2

    fragments:

    fragment run table:

    is update: false

    qualities:

    entries:

    fragment = 1, timestamp = 0, duration = 6000

    fragment = 2, timestamp = 6006, duration = 3000

    fragment: 3, 833788

    afra:

    long ids: true

    long offsets: true

    entries:

    time = 9009, offset = 8419

    global entries:

    bootstrap:

    version: 2

    is update: false

    is live: false

    movie id:

    timescale: 1000

    current media time: 12012

    smpte timecode offset: 0

    servers:

    qualities:

    segments:

    segment run table:

    is update: false

    qualities:

    entries:

    segment = 1, fragments = 3

     

     

     

    If you compare this part with that of what you have, you can observe that under fragment 2, you have the entry segment= 1 and fragments = 3 which is wrong. (my output has segment=1 and fragments =2 , which is correct).

     

    So, can you confirm whats the command you used for packaging ? Or probably, try with the default packaging and see if you still get the mismatched entries in inspect fragments.

     

    What build of FMS/File packager are you using ? Probably if thats old, you may want to switch to the most recent release.

     

    Thank you !

     
    |
    Mark as:
  • Currently Being Moderated
    Feb 23, 2011 5:01 AM   in reply to electroteque1

    Good to know that you have the issue resolved !

    I am little not sure about getting the wrong response code (503 instead of 404) because the f4f module does nothing but to redirect the status (error or otherwise) to apache which makes the response. Almost all the times when the configuration was wrong and when we had path related mismatches, 404 is what is recieved. Do post your observations as you try more and we can nail down the real issue

     

    But, can you please send me the exact f4fpackager command that you used while packaging for which you got the inspect fragments that you attached? I would like to reproduce and get the same inspect fragments ouput (with errors) and see what's causing that.

     

    Thank you !

     
    |
    Mark as:
  • Currently Being Moderated
    Nov 27, 2011 5:08 PM   in reply to electroteque1

    electroteque I'm having the same issue

    Can you tell me how you changed your configuration to work around it?

    I'm also wondering why the player is even asking for non-existent fragments ... I guess I need to get code to read the f4x for me ....

     
    |
    Mark as:

More Like This

  • Retrieving data ...

Bookmarked By (0)

Answers + Points = Status

  • 10 points awarded for Correct Answers
  • 5 points awarded for Helpful Answers
  • 10,000+ points
  • 1,001-10,000 points
  • 501-1,000 points
  • 5-500 points