golang中mysql建立连接超时时间timeout 测试

时间:2022-04-07 01:36:56

本文测试连接mysql的超时时间。

这里的“连接”是建立连接的意思。

连接mysql的超时时间是通过参数timeout设置的。

1.建立连接超时测试

下面例子中,设置连接超时时间为5s,读超时时间6s。
MySQL server IP是192.168.0.101,端口3306。

每3s执行一次SQL。

// simple.go

package main

import (
        "database/sql"
        "log"
        "time"

        _ "github.com/go-sql-driver/mysql"
)

var DB *sql.DB
var dataBase = "root:[email protected](192.168.0.101:3306)/?timeout=5s&readTimeout=6s"

func mysqlInit() {
        var err error
        DB, err = sql.Open("mysql", dataBase)
        if err != nil {
                log.Fatalln("open db fail:", err)
        }

        DB.SetMaxOpenConns(3)
        DB.SetMaxIdleConns(3)
}

func main() {
        mysqlInit()

        for {
                log.Println("start")
                execSql()
                time.Sleep(3*time.Second)
        }
}

func execSql() {
        var value int
        err := DB.QueryRow("select 1").Scan(&value)
        if err != nil {
                log.Println("query failed:", err)
                return
        }

        log.Println("value:", value)
}

启动程序:

go run simple.go

之后,接着使用iptables将所有发送到MySQL服务的数据包drop掉.

清除所有的过滤规则,防止干扰:

sudo iptables -F

设置drop策略:

sudo iptables -A OUTPUT -p tcp --dport 3306 -d 192.168.0.101 -j DROP

这样就会触发重试机制,最后超时。

output:

2019/10/27 18:34:52 start
2019/10/27 18:34:52 value: 1
2019/10/27 18:34:55 start
2019/10/27 18:34:55 value: 1
2019/10/27 18:34:58 start
2019/10/27 18:34:58 value: 1
2019/10/27 18:35:01 start
[mysql] 2019/10/27 18:35:07 packets.go:36: read tcp 192.168.0.104:54462->192.168.0.101:3306: i/o timeout
2019/10/27 18:35:07 query failed: invalid connection
2019/10/27 18:35:10 start
[mysql] 2019/10/27 18:35:15 driver.go:81: net.Error from Dial()': dial tcp 192.168.0.101:3306: i/o timeout
[mysql] 2019/10/27 18:35:20 driver.go:81: net.Error from Dial()': dial tcp 192.168.0.101:3306: i/o timeout
2019/10/27 18:35:20 query failed: driver: bad connection
2019/10/27 18:35:23 start
[mysql] 2019/10/27 18:35:28 driver.go:81: net.Error from Dial()': dial tcp 192.168.0.101:3306: i/o timeout
[mysql] 2019/10/27 18:35:33 driver.go:81: net.Error from Dial()': dial tcp 192.168.0.101:3306: i/o timeout
[mysql] 2019/10/27 18:35:38 driver.go:81: net.Error from Dial()': dial tcp 192.168.0.101:3306: i/o timeout
2019/10/27 18:35:38 query failed: driver: bad connection
2019/10/27 18:35:41 start
[mysql] 2019/10/27 18:35:46 driver.go:81: net.Error from Dial()': dial tcp 192.168.0.101:3306: i/o timeout
[mysql] 2019/10/27 18:35:51 driver.go:81: net.Error from Dial()': dial tcp 192.168.0.101:3306: i/o timeout
[mysql] 2019/10/27 18:35:56 driver.go:81: net.Error from Dial()': dial tcp 192.168.0.101:3306: i/o timeout
2019/10/27 18:35:56 query failed: driver: bad connection
2019/10/27 18:35:59 start
[mysql] 2019/10/27 18:36:04 driver.go:81: net.Error from Dial()': dial tcp 192.168.0.101:3306: i/o timeout
[mysql] 2019/10/27 18:36:09 driver.go:81: net.Error from Dial()': dial tcp 192.168.0.101:3306: i/o timeout
[mysql] 2019/10/27 18:36:14 driver.go:81: net.Error from Dial()': dial tcp 192.168.0.101:3306: i/o timeout
2019/10/27 18:36:14 query failed: driver: bad connection

从输出结果可以看到,首先打印读超时错误,6s超时(即readTimeout):

2019/10/27 18:35:01 start
[mysql] 2019/10/27 18:35:07 packets.go:36: read tcp 192.168.0.104:54462->192.168.0.101:3306: i/o timeout
2019/10/27 18:35:07 query failed: invalid connection

接着,每次执行SQL,都有3条连接错误日志,表示建立连接时,会尝试3次,每次超时时间5s(即设置的timeout)。

这里就引出一个问题,为什么每次执行SQL时,会进行3次建立连接尝试呢?

查一下QueryRow底层实现代码:

func (db *DB) QueryRow(query string, args ...interface{}) *Row {
    return db.QueryRowContext(context.Background(), query, args...)
}

func (db *DB) QueryRowContext(ctx context.Context, query string, args ...interface{}) *Row {
    rows, err := db.QueryContext(ctx, query, args...)
    return &Row{rows: rows, err: err}
}

// QueryContext executes a query that returns rows, typically a SELECT.
// The args are for any placeholder parameters in the query.
func (db *DB) QueryContext(ctx context.Context, query string, args ...interface{}) (*Rows, error) {
    var rows *Rows
    var err error
    for i := 0; i < maxBadConnRetries; i   {
        rows, err = db.query(ctx, query, args, cachedOrNewConn)
        if err != driver.ErrBadConn {
            break
        }
    }
    if err == driver.ErrBadConn {
        return db.query(ctx, query, args, alwaysNewConn)
    }
    return rows, err
}

其中,maxBadConnRetries定义为2:

// maxBadConnRetries is the number of maximum retries if the driver returns
// driver.ErrBadConn to signal a broken connection before forcing a new
// connection to be opened.
const maxBadConnRetries = 2

当err是driver.ErrBadConn时,会尝试3次进行连接。这就是为什么刚才的日志中有3次连接错误。

2.reject测试

另外,如果使用iptables将所有发送到MySQL服务的数据包reject掉:

sudo iptables -A OUTPUT -p tcp --dport 3306 -d 192.168.0.101 -j REJECT

reject会直接返回RESET操作,将连接重置,不会触发连接建立超时时间。

日志输出:

2019/10/27 21:39:55 start
2019/10/27 21:39:55 value: 1
2019/10/27 21:39:58 start
2019/10/27 21:39:58 value: 1
2019/10/27 21:40:01 start
2019/10/27 21:40:01 value: 1
2019/10/27 21:40:04 start
2019/10/27 21:40:04 value: 1
2019/10/27 21:40:07 start
[mysql] 2019/10/27 21:40:13 packets.go:36: read tcp 192.168.0.104:54536->192.168.0.101:3306: i/o timeout
2019/10/27 21:40:13 query failed: invalid connection
2019/10/27 21:40:16 start
2019/10/27 21:40:17 query failed: dial tcp 192.168.0.101:3306: connect: connection refused
2019/10/27 21:40:20 start
2019/10/27 21:40:21 query failed: dial tcp 192.168.0.101:3306: connect: connection refused
2019/10/27 21:40:24 start
2019/10/27 21:40:25 query failed: dial tcp 192.168.0.101:3306: connect: connection refused
2019/10/27 21:40:28 start
2019/10/27 21:40:29 query failed: dial tcp 192.168.0.101:3306: connect: connection refused
2019/10/27 21:40:32 start
2019/10/27 21:40:33 query failed: dial tcp 192.168.0.101:3306: connect: connection refused

从输出中可以看到,首先打印读超时错误,读超时readTimeout设置为6s:

[mysql] 2019/10/27 21:40:13 packets.go:36: read tcp 192.168.0.104:54536->192.168.0.101:3306: i/o timeout

后续的日志,都是连接被reset,无法连接的错误。

3.参考

golang database sql DSN (Data Source Name)中的timeout, readTimeout

go-sql-driver/mysql