Regions and Availability Zones

Regions

A region consists of several isolated and physically separated availability zones within a geographical area (usually a city). Surfercloud provides dozens of regions globally, covering China, North America, South America, Europe, Asia-Pacific region, South Africa, and the Middle East.

Availability Zones

An availability zone is a set of resource combinations that are physically and electrically isolated from each other. An availability zone may be designed by combining one or more machine rooms. The physical distance between multiple availability zones helps prevent problems caused by power outages, fires, or natural disasters such as typhoons and earthquakes. Thanks to availability zones, partitioned applications can easily achieve high availability. The fault impact range will be isolated within a single availability zone after reasonable partition design.

Latency

The “latency” perceived by users includes “application layer processing latency” and “network layer transmission and propagation latency”, where:

• Application layer processing latency is related to the load of your computing nodes
• Network layer transmission and propagation latency refers to the following formula

Network layer transmission and propagation latency = virtual network processing latency + host network card transmission latency + physical network propagation latency

• When two computing nodes are in the same availability zone, their network layer transmission and propagation latency is usually within 0.5ms
• When two computing nodes are in the same region but different availability zones, their network layer transmission and propagation latency is usually within 2ms
• The actual latency is based on the ping test data between the two computing nodes.