locked
Voice Resiliency Question RRS feed

  • Question

  • I'm in the process of setting up failover routes in case my central site w/ the collocated mediation servers goes down.  What I've done so far is setup the backup registrar and tested client re-direction successfully.

    I've also created the failover routes and PSTN usages and added them to the appropriate voice policies but here is where I'm confused.

    In that scenario users PSTN calling will failover to another mediation server AND another gateway correct? If so, how does a gateway in another site with another telco handle calls coming from numbers not local to the area?

    Is there any way to have it just failover to another mediation server? My confusion is you can only add a PSTN gateway to 1 Mediation server in the Topology builder.

    I hope I haven't come across too dumb here but this is really making my head hurt :)

    Thanks in advance for the help!!

    Thursday, April 21, 2011 1:33 PM

Answers

  • Hi

    You are right, you can only add the gateway to one medation server or medation pool. (You can sometimes trick Lync into point to the same gateway by using FQDN for one and IP for another, or alternate hostnames if your gateway supports it.)

    If you wish to failover to another medation server, you should set the medation servers up as a pool. I assume by what you have written you are using standard edition? In which case to setup a medation pool you will need to split out the medation role. In enterprise edition colocated medation servers will act as a pool.

    By default if you fail over to another gateway and medation server your outbound call will likely display a CLI from that line. For inbound you will need to talk to your telco about how they can handle redirecting the calls, this is different on a telco by telco basis.

    Hope this helps

    Tom


    Tom Arbuthnot, Consultant Modality Systems
    • Marked as answer by Diocletian Tuesday, April 26, 2011 4:47 PM
    Thursday, April 21, 2011 2:22 PM

All replies

  • Hi

    You are right, you can only add the gateway to one medation server or medation pool. (You can sometimes trick Lync into point to the same gateway by using FQDN for one and IP for another, or alternate hostnames if your gateway supports it.)

    If you wish to failover to another medation server, you should set the medation servers up as a pool. I assume by what you have written you are using standard edition? In which case to setup a medation pool you will need to split out the medation role. In enterprise edition colocated medation servers will act as a pool.

    By default if you fail over to another gateway and medation server your outbound call will likely display a CLI from that line. For inbound you will need to talk to your telco about how they can handle redirecting the calls, this is different on a telco by telco basis.

    Hope this helps

    Tom


    Tom Arbuthnot, Consultant Modality Systems
    • Marked as answer by Diocletian Tuesday, April 26, 2011 4:47 PM
    Thursday, April 21, 2011 2:22 PM
  • Thanks Tom!

    Here is the topology breakdown:

    2 EE FE servers w/ collocated Mediation server pool at central site

    1 SE FE backup registrar w/ collocated Mediation server

    currently 2 sites with their own mediation servers so they will be fine if central site WAN goes down since backup registrar has tested successfully for logins.

    5 other sites that rely on central site mediation server pool so what I would like is for those sites to failover to one of the mediation servers elsewhere but not to both the mediation and PSTN gateway offsite.

    Should I try to create the same PSTN gateway w/ FQDN and add it to a different mediation?  Is there any other way to accomplish?

    Thanks again

     

    Thursday, April 21, 2011 2:54 PM
  • Our gateways (OfficeMaster by Ferrari electronic) support SAN certificates so you can talk to them from different FE servers using different FQDN pointing to the same IP address - and still use TLS. The gateways can be in different areas using appropriate dialing rules - all numbers internally should be E.164.
    Johann Deutinger | MCITP Lync 2010 | MCTS Exchange 2010, OCS | ucblog.deutinger.de | http://twitter.com/jwdberlin
    Monday, April 25, 2011 11:50 AM
  • Hi again,

    What gateway are you using? Using TCP or TLS?

    When you say 2 sites with their own mediation servers, I presume you are talking about the central site with the Enterprise pool and a different site with the backup SE? or are you talking about two additional sites?

    Best option is to expand the mediation pool if possible rather than adding a separate mediation pool, but if you go the separate pool route you can try adding by an alternate FQDN or IP if using TCP.

    cheers

    Tom

     


    Tom Arbuthnot, Consultant Modality Systems
    Monday, April 25, 2011 3:09 PM
  • Hi Tom, you're suggestion to use FQDN to "trick" Lync worked perfectly.  I was able to setup failover routes/PSTN usages for the site voice policies and doing a failure test worked exactly the way i wanted.

     

    Thanks for the help and the info

    cheers

    Eric

    Tuesday, April 26, 2011 4:47 PM
  • Glad I could help

    cheers

    Tom



    If a post is helpful, please take a second to hit the green arrow on the left, thanks
    Tom Arbuthnot, Consultant Modality Systems Blog: Lync'd Up Blog


    Wednesday, April 27, 2011 9:05 AM