golang - go-sql-driver/mysql for rows.Next()的性能问题

Question

1.当查询出来的记录过万时，遍历Rows的时间过长，达到1分钟甚至更多

程序代码：

rows, err := p.conn.Query(sqlStr, params...)
if err != nil {
    log.Println("mysql query error", err.Error())
    return nil, err
}
log.Println("conn结束:", goutil.GetCurrentTime())
//延时关闭Rows
defer rows.Close()
//获取记录列
if columns, err := rows.Columns(); err != nil {
    return nil, err
} else {
    //拼接记录Map
    values := make([]sql.RawBytes, len(columns))
    scans := make([]interface{}, len(columns))

    for i := range values {
        scans[i] = &values[i]
    }
    //此处遍历在3W记录的时候，长达1分钟甚至更多
    for rows.Next() {
        _ = rows.Scan(scans...)
        each := map[string]interface{}{}
        for i, col := range values {
            each[columns[i]] = string(col)
        }
        rowMaps = append(rowMaps, each)
    }
    return rowMaps, nil
}

Answer 1

You can create multiple goroutines to query data in segments. Use channl to get the query structure in the thread for iteration.

Answer 2

This is not a problem of golang performance, other programming languages ​​will be similar. Fetching 30,000 records at one time is an unreasonable demand. It should be solved through paging and only fetching dozens to hundreds of records at a time.