Reliable

Speed of Response

We endeavour to reply to issues on the same working day, but if it’s a tricky problem then it might take us a little longer to answer your query fully. We can be contacted a variety of ways, but recommend that for the fastest response to urgent, service-affecting problems: please call us by phone.

Uptime

You can judge us by our uptime, and we will credit you back (as per our service level agreement) if we fall short of our promises to you. So that we are able to make this kind of commitment to you we do our best to find technical solutions to eliminate the risks that come from living in an imperfect world.

Hardware Reliability

Even the best made equipment can fail, and so we take a pragmatic approach to how we choose, configure, and operate our hardware:

  • dual power supplies (fed from separate PDUs, in turn fed from separate sources)
  • colour-coded power cables to make it obvious that everything is redundantly fed
  • hot-swappable parts where justified by our risk assessment matrix
  • real-time data replication of VPS volumes to separate physical hardware
  • a blend of traditional RAID, DRBD, and ZFS for disk arrays
  • periodic SMART testing of storage devices
  • “TBW” and wear-levelling monitoring of solid state storage
  • temperature and other environmental monitoring to track hotspots within our equipment

Network Reliability

Our network is multi-homed (that is to say we have multiple links between our network and “the rest of the Internet”) and spread across multiple sites. We interconnect these sites with physically diverse connectivity to mitigate the risk that a single point of failure could cause.

On top of this “designed-in” reliability, we observe the performance of our links continuously. To do this we examine our network — and our customers’ networks — around the clock, both from the inside and from external monitoring vantage points around the world. These feed back into two separate systems (Prometheus and Icinga2/Graphite) for anomaly detection and alerting.

Data-centres

We choose our data-centres to have facilities exceeding the Tier 3 standard, sometimes informally called “Tier 3+” within the industry:

  • dual feeds of power to customer racks
  • redundant uninterruptible power supplies
  • redundant power generation
  • redundant environmental controls
  • diverse redundant connectivity paths into the building
  • sufficient on-site generator fuel
  • periodic testing of the power generation equipment
  • fire detection
  • fire suppression

The data-centres we use are accredited to assure both the security and reliability of the services they provide:

  • ISO 9001 (quality management)
  • ISO 14001 (environmental management)
  • ISO 22301 (business continuity)
  • ISO 27001 (information security management)
  • PCI-DSS (Payment Card Industry Data Security Standard)