Mesh big data distribution
Cache processing is recommended. According to the amount of data you said, the cache based on redis can be fully satisfied, and the access speed can reach 10W+. In addition, whether the suggested hashMap is ConcurrentHashMap or other, whether the page display is incremental query or direct query again, and whether netty or mina is used for socket data reception need careful consideration and design. Such a large concurrent demand can be considered as a distributed cluster. It is estimated that this is exactly what the leaders want.