I have looked into this and I believe there is room for improvement (there always is). We should probably offer various strategies for reading media based on different factors. For instance reading a small file with NIO takes longer then reading it with standard IO in Java, but reading a large file with NIO is much faster then in standard IO.

Paul

On 10/30/06, Martin Schipper <[EMAIL PROTECTED]> wrote:
Hi,

I don't know exactly how the filestream mechanism works, but it seems to
scans the whole file before it starts serving.
We use Red5 for streaming quite large FLV files (200 ~ 700MB) and a
quite large archive (currently 120GB and growing every day). Those files
are kept on a fileserver.

Red5 scans the whole file (for metadata?) every time it's requested.
Isn't it possible to create some kind of cache for this procedure?
If this feature is implemented; Red5 is a serious competitor for M$
Streaming server :)

Kind regards,

Martin Schipper

_______________________________________________
Red5 mailing list
[email protected]
http://osflash.org/mailman/listinfo/red5_osflash.org



--
The early bird may get the worm, but the second mouse gets the cheese.
_______________________________________________
Red5 mailing list
[email protected]
http://osflash.org/mailman/listinfo/red5_osflash.org

Reply via email to