Understanding of the presence of a large number of close_wait states

Take the web server generating a large number of close_wait states as an example

1.close_wait status introduction:

The client actively closes the connection, the server receives the client's FIN, but has not sent its own FIN. The state at this time is the close_wait state, and a large number of close_wait states drag down the server performance

2. Reason for close_wait:

In some cases, the client closes the connection, but we are busy reading and writing, and the connection is not closed

3. Workaround:

Idea: Check out the connection that the client has closed, close him

The reason why this problem occurs must be that there is a problem with the code of the connection release on the server side

1. When the server fails to read and write, you can choose to close the connection

2. Periodically send inquiry data to the connection and check the received reply data packets (Heart-Beat thread sends heartbeat data packets in the specified format)

3. Modify the keep-live parameters (timeout time, tcp check interval time: the interval between keeplive detection packets sent, tcp check times: if the other party does not respond, the number of times the detection packet is sent)

ps: There are a lot of close_wait states on the server. This pot must be the back of the coder on the server side!