<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Hi Leo,<br>
<br>
The most recent material I can find cites under 20 msec one way for
h323; however this writeup is pre- inbound/outpbound faststart. For
MGCP you can be up to ~70 msec one way. Beyond that the total media
setup time exceeds 500 msec and becomes noticeable to the end user.
Contact your Cisco account team with an NDA to get additional details.<br>
<br>
/Wes<br>
<br>
Leonardo D'Urso wrote:
<blockquote
cite="mid:1083021520-1193343385-cardhu_decombobulator_blackberry.rim.net-1036260840-@bxe006.bisx.produk.on.blackberry"
type="cite">
<pre wrap="">Thanks Wes
I will wait your next email about this question.
Ciao
Leo
---
Leonardo D'Urso
<a class="moz-txt-link-abbreviated" href="mailto:l.durso@gmail.com">l.durso@gmail.com</a>
-----Original Message-----
From: Wes Sisk <a class="moz-txt-link-rfc2396E" href="mailto:wsisk@cisco.com"><wsisk@cisco.com></a>
Date: Thu, 25 Oct 2007 10:43:02
<a class="moz-txt-link-abbreviated" href="mailto:To:l.durso@gmail.com">To:l.durso@gmail.com</a>
Cc:ciscovoip <a class="moz-txt-link-rfc2396E" href="mailto:cisco-voip@puck.nether.net"><cisco-voip@puck.nether.net></a>
Subject: Re: [cisco-voip] Callmanager cluster in Wan
I've seen many deployments that use this model. Noteworthy items:
1.register a phone and gw in the US to the EU servers and test the
timing. With high RTT cm<->gw and cm<->phone users may notice a delay
in audio cut through when answering calls. Faststart for h323 and
optimized mgcp help some, but delays can still become an issue. I will
post a followup message with details when I locate them.
2. dialpan on EU servers will become significantly more complex hosting
PSTN patterns for US and EU
3. Consider SRST for the US site, possibly using multiple SRST routers
to allow all phones to register in the event of any network outgate
between US and EU.
4. if you deploy 3rd gen ipphones (7941/7961/7911/7970/7971, etc.) be
aware of upgrade issues due to TFTP over high RTT. Strongly consider
upgrading to CM version and phone versions that support peer to peer
firmware sharing. (cm6.0 srnd page 3-15)
More details of this are covered in the SRND:
<a class="moz-txt-link-freetext" href="http://www.cisco.com/go/srnd">http://www.cisco.com/go/srnd</a>
Your scenario is covered under "multi-site wan with centralized call
processing' but depends on information in other sectsions as well.
/Wes
Leonardo D'Urso wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Hi there
A customer of mine have two site: one in EU, one in USA with two different clusters of callmanager. Actually there is a trunk that permits to call between.
I have users that want to have the same extension mobility profile on each side.
I have evaluated that with a latency of 120ms it is not possible to have 1 centralized ccm cluster infrastructure with 1pub and 2sub in EU and 1 sub in USA. But what happens if i have all ccm server in a single location EU and a single voice gateway with SRST in USA and the ip phones that directly point to ccm cluster in EU?
Regards
Leo
---
Leonardo D'Urso
<a class="moz-txt-link-abbreviated" href="mailto:l.durso@gmail.com">l.durso@gmail.com</a>
_______________________________________________
cisco-voip mailing list
<a class="moz-txt-link-abbreviated" href="mailto:cisco-voip@puck.nether.net">cisco-voip@puck.nether.net</a>
<a class="moz-txt-link-freetext" href="https://puck.nether.net/mailman/listinfo/cisco-voip">https://puck.nether.net/mailman/listinfo/cisco-voip</a>
</pre>
</blockquote>
<pre wrap=""><!---->_______________________________________________
cisco-voip mailing list
<a class="moz-txt-link-abbreviated" href="mailto:cisco-voip@puck.nether.net">cisco-voip@puck.nether.net</a>
<a class="moz-txt-link-freetext" href="https://puck.nether.net/mailman/listinfo/cisco-voip">https://puck.nether.net/mailman/listinfo/cisco-voip</a>
</pre>
</blockquote>
</body>
</html>