[asterisk-users] Core show translation > 4000ms

[Tony Mountifield]

In article <4E85D19F.4090602 at digium.com>,
Kevin P. Fleming <kpfleming at digium.com> wrote:
> On 09/30/2011 07:49 AM, Administrator TOOTAI wrote:
> > Le 30/09/2011 14:05, Kevin P. Fleming a écrit :
> >> On 09/30/2011 03:56 AM, Administrator TOOTAI wrote:
> >>> Hi list,
> >>>
> >>> we have 2 asterisk boxes in VM (kvm) on 2 different Dell servers, one is
> >>> Lenny kernel 2.6.26 asterisk 1.6.2.20, the second CentOS 2.6.18 asterisk
> >>> 1.4.36 (Elastix). Both 64bits, no hardware involved, dahdi on both
> >>> machines for meetme timing.
> >>>
> >>> Doing core show translation give on the Lenny server
> >>>
> >>> Translation times between formats (in microseconds) for one second of
> >>> data
> >>> Source Format (Rows) Destination Format (Columns)
> >>>
> >>> g723 gsm ulaw alaw g726aal2 adpcm slin lpc10 g729 speex ilbc g726 g722
> >>> siren7 siren14 slin16
> >>> g723 - - - - - - - - - - - - - - - -
> >>> gsm - - 2 2 4001 2 1 2 - - - 4001 4002 - - 4003
> >>> ulaw - 4001 - 1 4001 2 1 2 - - - 4001 4002 - - 4003
> >>> alaw - 4001 1 - 4001 2 1 2 - - - 4001 4002 - - 4003
> >>> [...]
> >>>
> >>> and on the CentOS one
> >>>
> >>> g723 gsm ulaw alaw g726aal2 adpcm slin lpc10 g729 speex ilbc g726 g722
> >>> g723 - - - - - - - - - - - - -
> >>> gsm - - 2 2 2 2 1 3 - 6 - 2 2
> >>> ulaw - 2 - 1 2 2 1 3 - 6 - 2 2
> >>> alaw - 2 1 - 2 2 1 3 - 6 - 2 2
> >>> [...]
> >>>
> >>> Why do we have such latency on the Lenny machine for the codec
> >>> translation? Is this due to a kernel parameter?
> >>
> >> Because you didn't read the output. It clearly says "(in
> >> microseconds)" in the 1.6.x output.
> >>
> >
> > Well, I surely ask the wrong way, sorry: ms or us, 4001 from ulaw to gsm
> > and 2 the other way, still a huge difference. The output from centos
> > shows similar value in both directions.
> 
> This is why the output was changed to microseconds from milliseconds; in 
> the older version, the lowest number that should be shown was 1 
> millisecond, even if the actual amount of time consumed was 10 
> microseconds (or less). The "1" numbers in the output from the older 
> could easily have been "0.02", which would be closer to the output from 
> the new version.

Maybe, but that still doesn't explain why there is a factor of 2000
between some conversions and others. And 4001, 4002 and 4003 are
remarkably like a big round number plus a tiny offset! I would agree
with the OP that the values shown look suspicious and would bear
some investigating...

Cheers
Tony
-- 
Tony Mountifield
Work: tony at softins.co.uk - http://www.softins.co.uk
Play: tony at mountifield.org - http://tony.mountifield.org