Micron21’s Geographically Redundant IP Service is designed for the most mission critical hosting: whether you have a Co-Located, Physical Dedicated or Virtual Dedicated Server. Imagine - Two Virtual or Physical Servers existing in real time synchronization on independent, fully redundant hardware in two geographically separated datacentres but with the SAME public Facing IP address, providing a secure and stable high availability environment. This is the definition of our “Micron21 Geographically Separated Fault Tolerance IP Service”.
Normally services in different datacentres have difference IP addresses, causing issues when you start planning high availability services. The typical solution for redundancy or load balancing is to rely on DNS for failover. Unfortunately this method of redundancy has many common issues such as TTL (Time To Live) problems and incorrect IP addresses being cached on local DNS servers, resulting in extended downtime (not uptime) in the event of any service going off line.
Micron21 has solved this problem by being able to supply you the same public facing IP address in two independent geographical separated locations, which provides you with completely transparent redundancy and avoids all DNS issues. Our Geographical Redundant IP Services can be provided as a fully redundant Active – Standby service across two locations, or as a fully redundant and load balanced Active – Active service across two locations.
The technology driving our Micron21 geographical redundant IP service is based on a combination of OSPF, RIP and BGP routing, bringing the heart of Internet redundancy routing technology to individual client services. We can deploy this technology to an individual physical Ethernet port anywhere on the Micron21 Network right the way down to an individual Virtual Dedicated Server running any compatible operating system.
Micron21 can customize the time of the fail over response time the primary site to the secondary site from an almost instant fail over, to minutes or hours if required, or we can load balance each service in real time across both independent geographical separated locations with each location fully redundant from the other. The level of redundancy, response and high availability configuration is entirely up to you!
Geographically Separated Key Features
|Complete live backups daily, weekly and monthly of your Co-Location Server, without the need to shut down your image|
|Geographically separated backup storage|
|Synchronization data purchased either on the usage level, or mbit level|
|Any Operation System – Both 32 and 64 bit|
|Juniper Hardware Firewall Physical Port per customer
|Instant automatic fail over to the second location, upon the primary server failing|
|120 second or less fail over if the entire primary datacentre goes off air (limited to the speed of BGP routing)|
|Independent mains power with both UPS and Diesel generated power at each locations|
|Fully Independent redundant BGP multi homed Internet Connections at each location|
Detailed Technology Overview
Micron21 achieves this level of redundancy by allocating each High Availability Service 5 routable /29 (or /30) IP address allocations. (A1, A2, B1, B2 and C)
The Primary and the Secondary sites are located in independent datacentres, on independent networks with independent power on totally independent hardware.
The Primary site is allocated two /29 IP routable allocations A1 and A2 accessible fully independently from each other and the secondary location.
The Secondary site geographical located 45km away from the primary site is allocated two separate /29 IP routable allocations B1 and B2 accessible fully independently from the primary location.
IP Address Allocation C forms your public facing High Availability IP address which automatically floats via BGP and RIP routing between each physical location providing for a fully geographical redundant service.
In the event of a local issue on the physical hardware or even individual Virtual Dedicated Server, traffic within 5 seconds is re routed internally to the secondary site 45km away.
In the event of an upstream provider issue or even an entire datacentre issue at the primary site traffic is re routed automatically via BGP within 30 to 90 seconds.
There is no need to change name servers or DNS records the change is completely transparent and automatic.
Both the Primary and Secondary site can be accessed independently via fully independent connections at any time. Each service is kept in Synchronization via a dedicated network between IP address Allocation A2 and Allocation B2, configured with either manual or fully automatic real time Synchronization.
In the event of even more redundancy you can add to the mix features such as VMware vMotion or even Fault Tolerance.