California Solar Powered Data Centre Specification
Solar Powered Data Centre
Electricity from 120 solar panels (owned and operated by the data centre) capable of generating up to 60 kilowatt-hours of electricity per day provides all power for data centre offices, air conditioners and all computer equipment (network and servers).
DC power from the solar panel feed is stored in battery banks. A large APC UPS (Uninterruptible Power Supply) system also provides power backup and conditions the power supply protecting equipment from voltage fluctuations and spikes.
In an emergency, power is also available from natural gas standby generators (tested weekly), but in practice these have never been used.
Solar tubes have been installed eliminating the need for electric lighting during daylight hours, saving even more energy.
24h System Monitoring
The data centre monitor our servers each minute, 24 hours per day, 7 days per week using multiple independent redundant monitoring systems. Alerts are sent to data centre technicians immediately in the case of any equipment malfunction.
We also remotely monitor all servers every minute from the UK and multiple other locations around the planet. Any issues found and confirmed by several locations will generate alert messages.
Cisco Powered Network
Fully redundant Cisco powered network with multi homed connectivity to the Internet backbone: no single point of failure; redundant Cisco ASR 1001 series routers running full BGP routes, Cisco ASA 5525-X series firewalls and Cisco Catalyst 4500 series Gigabit switches.
The high availability network infrastructure is fully redundant with diverse leased fibre, dark fibre and FCC licensed wireless links with redundant route optimised BGP bandwidth peering with providers Internap, Level3, Hurricane Electric, Time Warner, Highwinds, and many others.
Redundant connections are connected to OC-192 fibre optic SONET rings that peer in Los Angeles, San Diego and Palm Springs. In our data centre we use redundant high-end Cisco routers and switches to connect us to the Internet. These routers are continually active and will automatically redistribute Internet traffic load using our own in-house Internap FCP optimisation hardware, resulting in minimum to zero downtime, lower latency, less jitter and packet loss and better performance from first mile through middle mile.
They chose Internap for IP as well as having Internap equipment on site because of its performance and reliability (used by many other companies including Microsoft and Sony). They buy directly from Internap and Internap gets its high performance by buying direct connections to many Tier 1 providers, including UUnet (Verizon/alter.net), AT&T, Sprint, XO, Savvis, Verio/NTT, Global Crossing and more. With Internap Performance IP, you don’t need to worry about your data traveling around to different backbones to reach a peering point, it gets put on the right network immediately, right at the source. There is no middleman, no shuffling between backbones. That translates to fewer hops and lower latencies.
Data centre uptime guarantee to Athenaeum (Ecological Hosting) is 99.9+%.
VMware ESX Platform with VMotion
The data centre operates VMware virtual server technology powered by multi-processor IBM BladeCenter servers with large amounts of onboard memory (RAM) and multi-terabyte RAID protected Fibre Channel SANs (storage area networks) from NetApp.
Compared to large numbers of individual physical servers this solution provides massive savings in data centre power consumption as well as noise and other environmental factors. In combination with the SAN and specialist software, each of the new physical servers is capable of running multiple virtual servers through the use of machine hardware emulation. These measures result in increased efficiency giving the ability to make more use of the resources available and generate considerable savings in energy consumption as well as reducing the overall amount of equipment required.
VMware VMotion brings a whole new level of redundancy to our server setup. If a physical server is struggling under the load or in the unlikely event of major hardware failure, VMware virtual servers hosted by the failing/failed physical server will automatically be switched to run on a different physical host. This is made possible by virtue of the servers being virtual and having all the server information stored away from the physical host in the SAN storage arrays.
The SAN also brings additional benefits: Multiple snapshots are taken of each virtual server stored on the SAN throughout the day. This enables server restoration to an earlier time almost instantaneously if needed. These snapshots are created by the SAN hardware and are completely independent of the servers.
Athenaeum (Ecological Hosting) has full remote control over our servers, including remote console access (i.e. even if the server is effectively off the Internet for some reason, we still have access as if we were sat in front of the computer), the ability to power cycle servers and server resource usage monitoring capabilities.
- Tier 3+ data centre
- Redundant, Multi-Homed Optimised Internap FCP Bandwidth
- Redundant Cisco ASR 1001 series routers running full BGP routes
- Redundant Cisco ASA 5525-X series firewalls
- Redundant SOURCEfire Snort 3D Intrusion Systems (IPS/IDS)
- Network Wide DoS/DDoS automated attack mitigation
- Network Wide IP Reputation Threat Defence to block criminal IPs
- Redundant Cisco Catalyst 4500 series Gigabit switches
- Redundant N+1 Power and Cooling
- Redundant Network Cards in all servers
- Redundant Power Supplies in all servers
- Redundant AMD powered servers
- Redundant Clustered NetApp SAN (storage area network)
- Redundant Backup Servers
- Redundant, Fault Tolerant Web & E-mail Servers running in VMware
In addition to the server disaster recovery snapshots, overnight processes backup the data on each server to a remote location over secured channels and onto encrypted storage.
Important note: clients are responsible for their data and should keep full copies of their web sites and any associated data (e.g. databases), safe on their own systems.