Introducing: Five Nines
A fault tolerant, self healing, auto scaling hosting service for mission critical business applications.
Fully replicated in multiple data centres, each running on its own physically distinct, independent infrastructure.
Self aware server cluster always keeping a healthy quorum by automatically detecting and replacing faulty nodes within minutes.
Quickly responds to increased user demand by automatically growing the nodes of the server cluster and spreading the work load.
Web Hosting vs High Availability
Hardware failure is the most common contributor to unexpected downtime.
Standard Web Hosting
Simple and low cost, but suffers from single point of failure flaws which increases risk of downtime and crippling recovery times.
High Availability Hosting
Sophisticated network of servers with constant health monitoring and automated failover mechanisms.
A blueprint for
highly durable applications.
Traffic Load Balancing
Provides fault tolerance by distributing incoming requests to a cluster of web servers, read more
Web Server Cluster
Multi server web application environment spread across 3 data centres with self healing capabilities, read more
Database Server Cluster
Synchronous replication, automated failover and query distribution across multiple servers, read more
Traffic Load Balancing.
An Internet facing entry point that distributes incoming user requests to a cluster of web servers.
Load balancers are constantly aware of the state of the web server cluster and will automatically redistribute traffic whenever a web server fails or appears to be under stress.
Being the only Internet facing entry point to the hosting environment, load balancers form a hardened outside layer that shields the application code and data from any direct access.
The load balancing servers are distributed accross 3 different data centres. In the event of a sever network outage, traffic will automatically be routed to the other operational data centers while the fault is being attended to.
Web Server Cluster.
A multi server hosting environemnt distributed across 3 different data centres.
At any give time, application code is served concurrently from multiple redundent servers, which in turn are spread across multiple redundent data centres.
There is safety in numbers... A server or even an entire data centre failure can be tolerated by the cluster seamlessly without any adverse effects on the operation of the application.
In the event of a server failure, the faulty server is immediately removed from service and a new server is automatically added in the same data center within 5 - 10 minutes to replace it.
The Web Server Cluster is capable of handling rapid changes in network traffic patterns by automatically adding more servers to the cluster in times of high demand.
Database Server Cluster.
Synchronously replicated on multiple servers in multiple data centres with automatic failure built in.
Each database is synchronously replicated to a standby server in a different data center. Each data centre runs on its own physically distinct, independent infrastructure, and is engineered to be highly reliable.
In case of a server or an entire data centre failure, the availability impact of the application is limited to the time that the automatic failover takes to complete: typically under one minute.
Additional read replicas of the database is automatically added to the cluster during times of high demand allowing read query load to be spread across a larger number of servers.
Point in Time Recovery
Datbases are continuously backed up, which allows point-in-time recovery of your data. This allows us to restore your data to any second during the preceeding 14 days, up to the last five minutes.
The building blocks of
highly durable applications.
Customisable silos that provides each application its own operating system, libraries and configuration files.
Durable File Storage
Upon file upload, it is immediately stored on multiple devices across a minimum of 3 data centres.
In-memory Object Cache
Boost performance of database operation by retrieving data from a high throughput and low latency in-memory data store
Always available replica of your application with its own database for testing purposes.
Get In Touch
If you would like to get in touch about any of our projects, find us @fivesqrd or send a message below...