Redundancy and Recovery

Redundancy

Data Centre Level

Within every Australian data centre CloudCentral implements the following redundancy strategies:

• N+1 redundant Internet services from Australian TierP1 network service providers
• N+1 redundant Border Routers
• N+1 redundant Core Routers
• N+1 redundant 10PGigabit Storage Network Fabric
• N+1 redundant 10PGigabit Core Switches
• N+1 redundant 10PGigabit Distribution Switches
• N+1 bonded 10PGigabit network connections for each Storage, Compute and Management Host
• ZFS Mirrored + Striped configuration with multiple hot spare drives for Primary Storage
• ZFS RAIDZ2 dualPparity configuration for Backup Storage
• N+1 redundant Management Services
• N+1 redundant UPS, Generators, and Power Distribution
• Redundant Power Supplies on all Servers, Storage and Network devices
• Enterprise Grade Serial Attached SCSI (SAS) hard disk drives
• Enterprise Grade Solid State Drives (SSD)

Cross Data Centre / Regions

Across its multiple Australian data centres CloudCentral offers the following redundancy strategies:

• Australia-wide Load Balancing
• Application Replication leveraging native application specific replication technologies
• Database Replication leveraging native database specific replication technologies
• Backup Replication
• Storage Replication
• Redundant DNS services
• Protected path interPcapital Ethernet services

Disaster Recovery

CloudCentral Managed Enterprise Disaster Recovery is a managed disaster recovery service for protecting critical infrastructure.

This service works by installing a disaster recovery software agent into each physical or virtual server that requires protection.

There are two service levels catering for different recovery time objectives: Hot and Cold service levels.

Hot Disaster Recovery

The first tier service level provides for ‘Hot’ DR, which provides a one-for-one replica server hosted in the data centre for each protected server. Changes from the protected server in the customer environment are replicated asynconrously in near real-time to the DR replica server located in the data center. In the event of a disaster, upon customer request the DR replica server can be made active within 15 minutes of receiving the request.

Cold Disaster Recovery

Cold DR provides a many-to-one replication of protected servers to a central DR repository hosted within a CloudCentral managed data center.

A DR agent is installed into each protected server, which then replicates changes from the protected server to the DR repository asynchronously in near real-time.

In the event of a disaster, upon customer request, a new virtual server will be provisioned at the data center, and the contents of the protected server will be restored to the newly provisioned server from the DR repository within 2 hours.