* You are viewing Posts Tagged ‘cpu load’

Types of bridging

Earlier I was using asterisk 1.4 and 1.6. In these version it used to
do native bridging and the CPU load was not very high. Now after
switching to asterisk 1.8 it has started to do remote bridging and the
CPU load has often started to peak.

Could this be a configuration issue. I have done the same SIP settings
that was earlier there in 1.4 and 1.6. I have ‘directmedia=yes’ and
‘directrtpsetup=yes’ in sip.conf and both the peers use the same
codecs and there are no nat issues as well

Please help

On Thu, Mar 29, 2012 at 7:29 PM, Phil Frost wrote:
> On Mar 29, 2012, at 08:43 , Deepesh D wrote:
>> What are the different type of bridging used by asterisk in a SIP
>> call? What is the difference between Packet2Packet bridging, Remote
>> bridging and Native bridging?
>
> Packet2Packet bridging is when RTP datagrams are forwarded by Asterisk without modification. This imposes little load on the CPU. Obviously this can only happen if both ends are using the same codec, and likely there are likely other less obvious conditions that must be met.
>
> Remote bridging happens when Asterisk can direct both ends to send media (RTP probably) to each other directly, by a SIP reINVITE, for example. Only works if both ends have a route to each other, Asterisk is configured to do it, each end shares a codec, and probably a dozen other more subtle conditions are true. In this case there is no load on Asterisk as it’s not even in the media path. It also means it can’t do things like intercept and act on DTMF or monitor the call.
>
> Native bridging is when media is forwarded with Asterisk, but for whatever reason (different codecs, maybe) Asterisk must inspect or modify the stream. Could mean a significant CPU load.
> –
> Phil Frost
> Macprofessionals
> office 248-893-0738
> direct 248-662-0809
>
>
>
>
> –
> _____________________________________________________________________
> — Bandwidth and Colocation Provided by http://www.api-digital.com
> New to Asterisk? Join us for a live introductory webinar every Thurs:
>               http://www.asterisk.org/hello
>
> asterisk-users mailing list
> To UNSUBSCRIBE or update options visit:
>   http://lists.digium.com/mailman/listinfo/asterisk-users

progressinband, how much extra CPU load?

Hi everyone,

We have an Asterisk 1.4.17 user who has problems with sometimes not getting
a ring tone on the calling phone.

We’re considering setting progressinband = yes, but would like to know how
much extra CPU load this will require? If anyone can give something even
roughly specific (eg “30% increase”) that would be great, rather than just
“lots”.

Also, are there any ATAs which are known to not work with progressinband =
yes? We have Polycom, Linksys and Audiocode.

Thanks for any advice,

Call recording format

What format are the actual calls in? Are they in G.711u/a format or
are they in something else (perhaps gsm?) format? I’m asking to find
out if Asterisk would need to transcode them.

On Mon, Nov 22, 2010 at 6:47 AM, Vilius Adamkavicius
wrote:
> Hi All,
> We have a requirement to record over 60 simultaneous calls. Our recording
> facilities are implemented using Monitor() over AMI. The thing we have
> noticed that making 60 simultaneous call recordings using wav CPU load is
> significantly higher (around 2 times more) than using gsm. Even writing call
> recordings to /dev/null makes a big difference in CPU load.
> What could be the reason for this? Is Asterisk updating wav headers every
> time it writes?
> What would be recommended hardware setup for over 60 simultaneous call
> records?
> Regards,
> Vilius.
>
>
>
> –
> _____________________________________________________________________
> — Bandwidth and Colocation Provided by http://www.api-digital.com
> New to Asterisk? Join us for a live introductory webinar every Thurs:
>               http://www.asterisk.org/hello
>
> asterisk-users mailing list
> To UNSUBSCRIBE or update options visit:
>   http://lists.digium.com/mailman/listinfo/asterisk-users
>

Asterisk Playback sound dropping on linphone

Hi,
I dial on A* from a linphonec to a Playback() extension, then suddenly
the sound stops after a while, without any notice.
I enabled debug both in linphone and A*, and the RTP packets are sent
from A* and received from linphone. It doesn’t matter whether I choose
alaw, ulaw, gsm as codec (besides changing cpu load of course).

How can I debug it? I’m using A* 1.6.2 and both linphone 2.x and 3.x.

I just need a console scriptable softphone, so maybe there’s an
alternative to linphone (which seemed good enough anyway!)…

Thank you,
Matteo

MixMonitor

Hi,
Have you noticed a marked increase in CPU load when using MixMonitor?

I use PHPAgi and Asterisk 1.6.2.9-2.

Mickael.