which mysqli functions do a client server round trip

Question

I want to understand how many client sever calls are made for a typical mysqli query?

Step(1) $result = mysqli_query($link, $query);

Depending on the type of query, we use other mysqli function after this like

mysqli_fetch_fields, mysqli_affected_rows, mysqli_insert_id, mysqli_fetch_row

etc. then we close the result object.

Now, is all data retrieved and stored in php memory after step (1)? Or mysqli_fetch_fields, mysqli_insert_id etc makes another call to mysql server?

Reason for asking: Trying to understand how mysqli calls work. But can not find such explanation anywhere for beginners like me.

Answer 1

PHP MySQLi API is built on MySQL C API. So it would be better if you have knowlegdes of it.

Basically, SELECT query could generate large ResultSet and this ResultSet is transfered from Server to Client when you call PHP's mysqli_store_result() (In C API, mysql_store_result()).

• C API mysql_fetch_row() just returns a pointer to MYSQL_RES* (which is already stored in PHP right after mysql_store_result(). But 'mysqli_fetch_row()` would require some memories to make PHP's array.

• mysqli_insert_id() (which is last_insert_id() of C API) just returns insert id of MYSQL connection data stucture which means there is no extra memory for insert id.

If you want to know how MySQLi works, I would recommand to learn MySQL C API and see PHP source codes.

Answer 2

mysqli_query runs the query on the server and returns false is the query failed, true is the query was successful but did not return anything (UPDATE query for example) or a mysqli_result otherwise. That mysqli_result is a class that extends Traversable interface, so yes, it's in memory. All other functions mysqli_fetch_fields, mysqli_affected_rows etc. are just methods in that class so those just read what's already in memory.

For more details, read this: php documentation