Avaya Redundancy Procedures

5 Elements to consider when implementing redundancy

By the end of this post ‘Avaya Basic Redundancy Procedures’ you will understand the different techniques to setup redundancy for your Avaya Aura or IP Office solutions, IP Phones, Media Servers, Uninterruptible Power Supplies, Network Switches, and Trunk failover. 

Not so long ago I was helping a customer to test redundancy for their Voice Solution which consisted on multiple branch sites, each site contained an LSP for survivability as well as local ISDN Trunks for local and backup in the event the link failure.

IP Phones were part of the customer solution, pooling DHCP from the Core site, but no backup DHCP was designed, in the event of the IP Phones’ lease expiring. I also found some other issues related to the topics listed below.

These are the 5 elements to consider when implementing redundancy=

• 1.- IP Phones
• 2.- Media Servers
• 3.- Uninterruptible Power Supplies
• 4.- Network
• 5.- Carrier
• *- resources

 

1.- IP Phones

When setting up IP Phones redundancy you need consider the phone type, DHCP Server Location, and Call Servers.

Phone Type – Some Avaya Phones use Option 242, and some use Option 176 to push the Call Server, File Server, and other information out the phones. Here you can specify the failover sequence for each registration request. Allowing to tell the endpoint where to register in the event of failure.

DHCP Server – In a multisite with multiple Communication Managers, it is important to have a backup DHCP Server supplying IP Addresses. Another consideration can be the 80/20 rule, which allows 80 per cent of the lease address to be delivered by the main DHCP Server, then the second DHCP Server will be responsible of the rest 20% left.

Lease duration – Normally an address will be assigned to a node (IP Phone) for about 8 days, you can maximize lease up to 24 days. By increasing the lease time you provide more time for the branch site to comeback online; giving the endpoints more time between IP Address renewal or lease request.

Location – If possible have these DHCP Servers located in different buildings.

2.- Media Servers

Avaya Aura – High Availability or HA which means ‘Cold Standby’ based on Avaya standards allow up to 20 minutes for the redundant or Non-Preferred Server come online, also known as Stand-by Server.

The downside to having a HA environment is the distance between the servers. Make sure to implement redundancy in the Network and Power distribution for both servers, always keep them in physically connected to different power sources if possible.

Caveats

*.- The initial data transfer from the Preferred to the Non-Preferred server might take up to two hours depending of the size of the database.

*.- 100 meters is the maximum distance between the servers

*.- The Non-preferred server only gets Dom-0 and C-Dom installed

*.- Once the redundancy takes place between both servers, all the VMs shuts down the Preferred or Active server to avoid IP Address duplication.

*.- License will be valid at both servers regardless the difference in MAC Addresses.

IP Office Resiliency – The same as the Avaya Aura Servers the IP Office offers a similar failover. No IP End-point licenses are needed to accommodate the existing IP Phones in the Redundant Location, only a Preferred Licenses will be required.

9.1 Resiliency Features

Location Based Resiliency – This feature allows IP Endpoints to re-register to a backup remote IP Office system based on location.

Automatic Fallback Recovery – Once the Primary IP Office Server comes back online, the Endpoints will comeback to their mother after 10 minutes.

IP DECT Line Resiliency – Allows the Wireless IP DECT phones to register to a remote IP-DECT Server through the IP Line.

Voicemail Pro Redundancy – A distributed voicemail can be added on a remote location for voicemail continuation.

Caveats

*.- VM Distributed Configuration – Only able to store mailbox messages, then it will utilize SMTP services to transfer the existing mailbox messages to the Primary Server.

*.- The Redundant IP Office needs to have the User Rights Form configured at both systems.

*.- No IP Endpoint license is required at the Redundant System.

3.- Uninterruptible Power Supplies (UPS)

The UPS Redundancy or N+1 UPS is also called parallel redundancy, similar to the Avaya Servers, it uses a backup Modules for failover. As a good rule of thumb keep the power sources separate, If possible from different power generators. Refer to resources for more details.

Caveats

*.- If both Primary and Secondary UPS Modules fail, it will transfer to battery power, and run until the battery or shutoff timer expires.

4.- Network

If Ethernet Redundancy has been configured in the Avaya Media Gateways, you can add another level of redundancy by adding ‘Network Redundancy’. Let’s say that the Primary Network Stack A goes down, then the redundant Ethernet port corresponding to the Media Gateway will take over and communicate through the Secondary Network Stack B. It is worth mentioning that there are a variety of Redundancy Protocols available. They are called FHRP (First Hop Redundancy Protocols).

Caveats

*.- L2 Loops (if a L2 Switch dies, then loops will flood the Switch, causing degradation or failure to the rest of L2 Network Topology. For this always implement STP or RSTP. See reference for more details.

5.- Carrier

For trunking it is always important to have Carrier Redirect as part of your design strategy. If a Branch SIP/PRI Trunk goes down, you will need the Service Provider (Carrier) to redirect those calls to another SIP/PRI Trunk Group, this will optimize the user’s experience, and it is a most when designing Townships and First Responders Solutions.

Caveats

*.- Transfer Delay – In my experience there was a 10 minutes delay, before calls were landing at the backup or redundant site. The caller was getting busied signals while the switchover was happening. Work closely with the Service Provider to test any failover conditions.

Questions

What has been your experience when testing redundant systems?

*.-Resources

Liebert  – Balancing, Scalability, and Reliability

Cisco Data Center Design Overview

FHRP