CPC H04L 47/2416 (2013.01) [H04L 47/32 (2013.01); H04L 65/40 (2013.01)] | 18 Claims |
1. An application instance deployment method, comprising:
receiving, by a global management platform, quality of service (QOS) requirement information from a first client, wherein the QoS requirement information comprises a first delay requirement, a second delay requirement, and a first quantity of connections, wherein the first delay requirement represents a maximum accepted delay, and the second delay requirement represents an expected delay, and the second delay requirement represents a smaller delay than the first delay requirement;
selecting, by the global management platform, a first available site that meets the first delay requirement, wherein the first available site is selected from among sites managed by the global management platform;
deploying, by the global management platform, one or more first application instances on the first available site, wherein a quantity of connections that can be established to the one or more first application instances is less than or equal to the first quantity of connections;
determining a first total communication delay of accessing all the application instances by the first client;
determining a second total communication delay of accessing all the application instances by the first client after a second application instance is moved;
comparing the first total communication delay and the second total communication delay: and
in response to determining that the second total communication delay is smaller than the first total communication delay, changing a deployment location of the second application instance according to a QoS information table.
|