MySQL Optimization: from more than ten seconds to 300 milliseconds

         Today's article still records the problem-solving process . When I tested the code yesterday , It takes about ten seconds to find a function . I just look at my own code logic , After a complete look at the code , According to the feeling, there are two reasons for the long time ： One is mysql Writing problems , Another possible problem is the circular query of the database . The code dream will not be uploaded , It may involve the privacy of the company

         The bottom is glued out sql part , You can see this sql all 7 More than seconds , The second picture is using explain Results after actuator , You can see that only two of the five tables are indexed , Other tables do not have indexes , You can see type The level of this column is also ALL.
         The result value is from good to bad ：system > const > eq_ref > ref > fulltext > ref_or_null > index_merge > unique_subquery > index_subquery > range > index > ALL. Generally speaking , Make sure that the query is at least range Level , It's better to achieve ref, Otherwise, there may be performance problems . So this sql To be optimized .

         How to optimize ：
        1, First combine type+key These two columns are used to optimize , That is, by increasing the index to optimize . I usually give it to where and order by Clause to add an index to a field . So I was there d Four indexes have been added to the table （deviceId、name、del_flag、project_id）,c Table to prdfession Also added indexes .

        2, Then according to Extra This column is optimized , You can see that the two columns in the above figure appear “Using temporary”、“Using filesort” In both cases .
        “Using temporary”： refer to sql If order by perhaps group by The fields in are from other tables instead of the first table in the join order , A temporary table will be created . Solution ： You can try to create a federated index .
        “Using filesort”： refer to order by The amount of field data queried exceeds mysql The default value of , Led to order by The fields in do not go through the index , File query is used , It is also the cause of performance bottlenecks . Solution ： You can try to put the sorting function into the code ,sql Do not sort in .
         When a normal index is created 、 Joint index （project_id + name） after , Look again. sql Execution speed of .

         It can be seen that the execution speed is immediately increased ,explain When the command is executed , Also will find type A whole train has reached ref Level ,extra There is no train left “Using temporary”、“Using filesort” These two signs that affect performance .
         Now? sql The optimization is complete , If the speed is still slow , Consider optimizing the logic of your code .