Thank you Romain.
I will try tomorow with an exit and with beta2.
And then i will stop using telnet and try to learn on_metadata :-)
In beta 1, I don't have the same logs. I have :
2011/07/19 02:01:41 [server:3] New client: localhost.localdomain
But immediatly followed by : Client left.
Best,
Alain
--
http://www.netvibes.com/bolli#Contact
2011/7/19 Romain Beauxis <[email protected]>
> 2011/7/18 Romain Beauxis <[email protected]>:
> > 2011/7/17 Alain Bolli <[email protected]>:
> >> Hi,
> >
> > Hi!
>
> Hi again!
>
> >> I don't know what on_metadata is. I'm a beginner with liquidsoap and I
> >> discover something new each day :-)
> >> I will look on savonet's website.
> >>
> >> I join the php file I'm using for retrieving the metadatas.
> >
> > Thanks for the file! As I suspected, you do not close the conversation
> > properly by sending a "exit" command but simply close the file
> > descriptor.
> >
> > Of course, the server should close its socket when you close the file
> > descriptor. There is clearly a bug somewhere here that I will try to
> > find.
>
> After testing a stripped-down version of your script -- :), I can
> reproduce part of what you describe. I see tons of:
> 2011/07/18 18:34:32 [server:3] Broken pipe in write()
> 2011/07/18 18:34:32 [server:3] Broken pipe in write()
> 2011/07/18 18:34:32 [server:3] New client: localhost
> 2011/07/18 18:34:32 [server:3] Connection reset by peer in read()
>
> Those lines are to be expected as they report the issue on
> liquidsoap's side when your client closes the connection without
> exiting properly. However, I have not been able to observe the other
> issue.
>
> The other issue is raised because too many connection are already
> opened and thus the server is not able to accept a new one.
>
> Finally, I am not sure about the status of this issue. The errors you
> see and report were actually not displayed in beta1 so it is very
> likely that they are actually happening in beta1 too.
>
> In all reality, I think that this is mostly a mistake on your client
> side: you should not poll too intensively the telnet server and also
> properly close the connection (send "exit" and maybe wait for the
> server to close on its side..).
>
> Thus, I do not think we are in the presence of a bug right now but
> I'll be ok to reconsider this if the issue shows up in a more regular
> case and is reproducible.
>
> In the mean time, I invite you to consider another implementation,
> such as for instance using an intermediary metadata store, updated
> through on_metadata.
>
> Hope this helps,
> Romain
>
------------------------------------------------------------------------------
Magic Quadrant for Content-Aware Data Loss Prevention
Research study explores the data loss prevention market. Includes in-depth
analysis on the changes within the DLP market, and the criteria used to
evaluate the strengths and weaknesses of these DLP solutions.
http://www.accelacomm.com/jaw/sfnl/114/51385063/
_______________________________________________
Savonet-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/savonet-users
