[cisco-voip] Clustering over WAN - Design question

[Jeff Garvas]

It sounds to me like your problem is the frequent WAN failure more so than
the lack of a subscriber at the remote location.   How is it acceptable for
the data network to be down so often?

Can you resolve that or build in redundancy with an equal cost, possibly
diverse path?


On Fri, Oct 2, 2009 at 10:07 AM, Phil G <pgciscovoip at gmx.net> wrote:

> That means, if i have the following environment:
>
>
> SITE A                          SITE B
>
> Pub                               Sub2
>        -------  WAN  ------
> Sub1
>
>
> Pub, Sub1 and Sub2 are all running ccm.exe, ICCS-real-time traffic would be
> 1,5 Mbps from Sub2 to Pub plus 1,5 Mbps from Sub2 to Sub1. Plus additional
> 1,5 Mbps ICCS database traffic from Sub2 to Pub. That means 4,5 Mbps over
> WAN. Correct?
>
> If i add another SITE C with a Sub3, then there would be an additional 1,5
> Mbps between SITE C and SITE B plus the existing 4,5 Mbps between SITE A and
> B and another 4,5 Mbps between SITE A and SITE C?
>
>
> Another question: Do you have any experience how good the database
> replication will start again after a WAN-failure. Problem is that we have a
> remote site with SRST. Now there are frequently WAN-link-failures and the
> people are complaining about the feature-set in SRST. The idea is now, that
> we place a subscriber-server at this site and raise the bandwidth. Now with
> the given reliability of the WAN-link, i am fearing about other "symptoms"
> we will facing when the link is up again (broken replications or something
> like that). But also when there are other issues on the link like more
> packet loss.
>
>
>
> Wes Sisk wrote:
>
>> Correct, it is per server.  It is not "per site".
>> On Friday, October 02, 2009 4:03:53 AM, Phil G <pgciscovoip at gmx.net>
>> wrote:
>>
>>> Hi!
>>>
>>> I have read the SRND-section of clustering over WAN and i have a question
>>> about bandwidth requirements: SRND states, that you need 1,544 Mbps minimum
>>> for sites that are clustered over WAN for ICCS real-time traffic. It says
>>> "sites" not "servers".
>>>
>>> And you need additional 1,544 Mbps for every subsriber server remote to
>>> the publisher for database traffic. Here is says "server" not "sites".
>>>
>>> Why is the real-time traffic independant of the number of servers at a
>>> site? I thought, the real-time traffic is fully meshed between each server
>>> running ccm.exe, so if i have 2 remote subscibers at a site, i would assume
>>> that i would have 1,544 Mbps to every other server running ccm.exe. So in a
>>> 2 site environment with 2 subsriber at each site i would have 4
>>> ICCS-real-time-traffic flows, each with 1,544 Mbps. Is that correct?
>>>
>>> _______________________________________________
>>> cisco-voip mailing list
>>> cisco-voip at puck.nether.net
>>> https://puck.nether.net/mailman/listinfo/cisco-voip
>>>
>>
>>
> _______________________________________________
> cisco-voip mailing list
> cisco-voip at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-voip
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://puck.nether.net/pipermail/cisco-voip/attachments/20091002/bad49439/attachment.html>