A character set problem caused by MySQL Skip-character-set-client-handshake _mysql

To help colleagues deal with a tricky business today, the problem is this:

Regardless of which version of the MySQL client the client connects to the server, it is found that as long as the server side is set

Copy Code code as follows:

Character-set-server = UTF8

After

Copy Code code as follows:

Character_set_client, Character_set_connection, Character_set_results

will always be consistent with the server side, even if the MySQL client plus options

Copy Code code as follows:

--default-character-set=utf8

No, not until you connect, and then manually execute the command.

Copy Code code as follows:

Set Names Latin1

, will be the client, connection, results character set to change.

After careful comparison, I finally found that the place where I went to the pit was that the server side set another option:

Copy Code code as follows:

Skip-character-set-client-handshake

The explanation for this option on the document is this:

Copy Code code as follows:

--character-set-client-handshake

Don ' t ignore character set information sent by the client. to ignore client information with the default server character set, use--skip-character-set-client-handshake; This is makes MySQL behave like MySQL 4.0

In this sense, it is also beneficial. For example, when you enable the Skip-character-set-client-handshake option, you can avoid client program misoperation, use other character sets to connect in and write data, which causes garbled problems.