Custard’s datacentre environment has been designed to ensure maximum uptime. Power is delivered via two separate grids from neighbouring towns and villages. Additional to the grid power, there is also power from solar energy from a large 60KW solar panel array. There is an automated diesel generator installed outside the datacentre which is thoroughly tested weekly. Additional to overall room power being very well protected, Custard have a UPS battery backup for each rack.
Custard have 3 x leased lines from 2 x separate providers. Two of the circuits are provided by BT and the 3rd circuit is provided by Virgin. Core switches and firewalls are redundant and we offer failover DNS services so that critical services such as RDP or email can be accessed, in the unlikely event of a line or firewall failure.
Custard’s datacentre facility has a member of staff onsite 24/7. The facility remains locked unless accessed by a member of staff with clearance. Cabinets are all coded and access to these are granted by senior members of staff to a handful of engineers with clearance to access racks for maintenance.
Custard’s WAN connections are protected by Next Generation Firewalls and web applications have the option of being protected by a Web Application Firewall. Datacentre Firewalls are only maintained by certified in-house engineers and we operate a change-tracking system with sophisticated reporting
Temperature is controlled by redundant air conditioning units. They run on a low power during normal circumstances and if one or more units fail the other units compensate to maintain the temperature in the room. Air-conditioning units are on different power phases and are protected by the redundant power systems including our diesel generator.
Custard have real-time monitoring of the infrastructure. If an alert is raised there are visual cues in the form of dashboard, red lights upstairs and downstairs and a ticket is automatically raised that alerts on-call engineers. Additionally, there is an audible alarm in the Network Operations Centre.
Virtual machines run from a cluster of latest generation hypervisors, these are monitored and patched constantly to proactively avoid disruption of services. Our hypervisors have redundancy of power built in but also fully redundant so If a hypervisor fails operations will continue. The virtual machines are stored on a central SAN with multiple hard drive and power failure tolerance built in.
Additional to the redundancy, protection and proactive work that Custard perform, all virtual machines are replicated to a separate cluster of the same specification so that they can be failed over to another cluster all together for Disaster Recovery. All data is backed up to a backup appliance which is replicated offsite to another backup appliance in a different geographical location, allowing for recovery in the event of a major disaster.