Is it possible to run a packet capture to see what happens over the wire
when this occurs? That would give some concrete information.
If you have the disk space, just rotate out a capture file every hour.
Once you catch it, filter out a pcap file with what happens when the
sources are disconnected, and send us a link.
Brad Isbell
brad at musatcha.com
http://www.musatcha.com
On Tue, Nov 4, 2014 at 5:58 AM, Dean Sauer <wemanageitfl at gmail.com>
wrote:
> On Fri, 31 Oct 2014 17:25:41 +0000, Dean Sauer wrote:
>
> > I have a IceCast Server as follows:
> >
> > 2048MB RAM 2048MB SWAP
> >
> > OpenVZ VPS Ubuntu 12.04 64b full update
> >
> > I've looked at logs and I am not seeing a reason for ALL THE FEEDS
to
> > DROP AT ONCE SAME TIME within milliseconds. Then they reconnect. Fine
> > for weeks or more, or might drop again... in a few seconds, minutes,
> > hours, or continue on for days and weeks...
> >
>
> The latest incident, yesterday AM, and of course I can't access it for
> the next 6 hours, so nothing but silence on those feeds.
>
> The server DROPS the SOURCES, BUT the SOURCES think they are connected.
>
> In this case, four feeds, think they are connected BUT the server has
> dropped them and is using the backup mount, which is to feed silence.
>
> The ONLY way to correct this is to drop the sources at the encoding
> source, and reconnect.
>
> In this case the sources are DarkIce 1.x via *buntu and Raspbian.
>
> It will happen on sources using non Linux based systems as well.
>
> Any hints at what to look for???? Other ideas???
>
>
> _______________________________________________
> Icecast mailing list
> Icecast at xiph.org
> http://lists.xiph.org/mailman/listinfo/icecast
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.xiph.org/pipermail/icecast/attachments/20141104/49299e16/attachment.htm>
