Nehos is an Australian Broadband Internet and Telecommunications Provider offering competitive, reliable and quality services to Business.
Our main website - [login to view URL]
We have been using OpenSIP’s (single instance) as the core of our voice infrastructure for the last 10 years along with Freeswitch gateways but the current setup does not have much redundancy.
We have attempted without success to upgrade our current setup to OpenSIPS 3.1 as a cluster for redundancy and load balancing.
Our current setup is single public IP address using a destination NAT rule on a Mikrotik Router to route SIP traffic to a single OpenSIPs instance.
All servers are running on Proxmox LXC containers. OpenSIPs is using a MariaDB cluster along with RedisDB for storing data.
We need to be able to fail over from one server to another for maintenance or emergency while keeping user location (SIP Registration) and SIP dialog information in sync between both servers. When we change the NAT to point the public IP from one node to the other we need inbound and outbound calls work reliably.
The current configuration is keeping this information in sync between both servers but when we try doing a test call we are having errors on the cli and the call fails.
What we need is someone to look at our current configuration and either fix it or make a recommendation on how we best achieve our goal.
The current setup is all within one location in our Brisbane data centre but we would also like to place an instance in our Sydney data centre so if we loose one location the other can take over. We would also like to know how we can setup load balancing similar to how DNS can route local traffic to the closest server.
As an additional side project we would like to have our current configuration looked at with suggestions on how you could improve and add to it?