The invention discloses a caching method applied to high-
concurrency requests, which comprises the following steps: 1) when a distribution platform service is started, reading data from a source
database (Redis) into a 2Q cache in advance; 2) when the management end service modifies the data, while updating the data to a source
database (Redis), pushing a copy of data to the RabbitMQ; 3) the distribution platform service subscribes to the RabbitMQ service, and when the data is updated, the corresponding
data information in the 2Q cache is updated according to the data pushed by the RabbitMQ; 4) when a
client initiates a large number of concurrent requests and requests to a distribution platform service, firstly judging whether the requests hit the 2Q cache, if so, executing the step 5), and if not, executing the step 6); (5) judging whether the 2Q cache is invalid, if so, executing the step (6), and if not, returning data to the
client; and 6) the
server side organizes and suppresses the request, and the organized and suppressed request is subjected to
data acquisition in a source
database (Redis). According to the method, service
paralysis caused by high-
concurrency requests can be effectively reduced, and meanwhile, the hardware cost caused by cache is reduced.