[Asterisk-Users] Any Gentoo Users Running ASTERISK had problems on recent emerge -u world?

[Vic Cross]

G'day David,

My * server runs Gentoo 1.4, and have recently started getting choppy
audio -- usually during voice prompts from voicemail and my talking clock
(digits dropped in the talking clock, and broken playback in voicemail:
"press star for help, and pound to exit", for example, comes out "press r
for help, and pound to-ound to ex-ound t-ound to exit".  I thought it
might have been a problem to do with a new NIC I installed a little while
back, but from your report I have to wonder!

The hardware is dual Athlon-MP 1600+, Tyan MB, 1GB RAM.  Asterisk is 
CVS-02/03/04 (hmmm, month old, might have to update soon).

On Mon, 1 Mar 2004, David Liu wrote:

> So I stop gracefully and started asterisk again.  It failed to start,
> complaining about music on hold.  With some further investigation, I
> confirm it is to do with MPG123.  I had to downgrade it to mpg123-0.59q
> or 0.59r in order to start Asterisk successfully again.

My mpg123 is still 0.59r-r3...  Thanks for the tip, I will not be updating 
this...

> Everything used to work until emerge -u world........perhaps next time I
> shouldn't really do that until I test i on a test server....oh well...

I have not done 'emerge world' recently, but have emerged a couple of
things -- nothing that is associated with the running of *, or is even in
use at present (there is an update to gcc and glibc available for example,
but I have not taken them yet).  On your machine, check out 
/var/log/emerge.log to see what packages got updated or installed after 
your 'emerge world'.  There might be clues in there.

I have noticed that the choppiness seems to occur when there is some
contention for CPU resource, although I have not been able to work out
what * is fighting against (load average doesn't change, nothing's getting
logged).  I'm chasing an odd occasional kernel oops on this machine also,
may be related somehow...  As this machine does all the local work for my
office, I've been thinking about putting the * workload onto its own
machine anyway -- I just don't trust Gentoo on the server anymore for some
reason :-(

What kernel is your machine running?  I updated to 2.4.25_pre6-gss a
little while ago, but I *thought* things were running okay after that.  
Prior to the current kernel I was on 2.4.23_pre8-gss-r2.

Sorry for not having a clear answer, but hopefully some thoughts to help 
us work out what might be going on.

Cheers,
Vic Cross