creating a copy of db using `mysqldump` command [closed] - mysql

Question

Closed. This question does not meet Stack Overflow guidelines. It is not currently accepting answers.
This question does not appear to be about programming within the scope defined in the help center.
Closed 20 days ago.
Improve this question
Is there a way to create a copy of a database with sample rows using mysqldump command?
myslqdump -u <username> -h <host> -p <database name> [<table name> ...]
I have a fairly large DB and need to create a copy so that a developer can work with the App. Instead of dumping the entire DB, is there a way to randomly sample rows and create a copy of the db?

Answer 1

Mysqldump does support row-level backup, as it has --where option which filters rows to be dumped.Here is what the reference manual says about it: --where='where_condition', -w 'where_condition'Dump only rows selected by the given WHERE condition. Quotes around the condition are mandatory if it contains spaces or other characters that are special to your command interpreter. But it might not be that user-friendly. Even if we put a subquery in a WHERE clause, we are still facing some restrictions which might not be possible to overcome. For instance, let's use actor table from DB sakila. It's legit to execute this in mysql cli:
select * from actor
where actor_id in (select *
from (select actor_id from actor order by rand() limit 5) t
);
+----------+-------------+-----------+---------------------+
| actor_id | first_name | last_name | last_update |
+----------+-------------+-----------+---------------------+
| 19 | BOB | FAWCETT | 2006-02-15 04:34:33 |
| 91 | CHRISTOPHER | BERRY | 2006-02-15 04:34:33 |
| 11 | ZERO | CAGE | 2006-02-15 04:34:33 |
| 120 | PENELOPE | MONROE | 2006-02-15 04:34:33 |
| 109 | SYLVESTER | DERN | 2006-02-15 04:34:33 |
+----------+-------------+-----------+---------------------+
However, it's erroneous to use the same WHERE clause when using mysqldump.
mysqldump -uroot -p sakila actor --where="actor_id in (select * from (select actor_id from actor order by rand() limit 5) t )" > /tmp/acto
r_bck.sql
-- error message:
mysqldump: Couldn't execute 'SELECT /*!40001 SQL_NO_CACHE */ * FROM `actor` WHERE actor_id in (select * from (select actor_id from actor order by rand() limit 5) t )': Table 'actor' was not locked with LOCK TABLES (1100)
Besides, as Shadow stated, retaining referential integrity is an issue when using mysqldump. We don't want a broken relationship between tables and unusable dataset. With all regards, please do not use mysqldump for random row-level sampling.
Under the circumstances, the best I can come up with is to use a stored procedure to do the row-level backup to a new database,with contents like:
create database sakila_bck
create table sakila_bck.actor select * from sakila.actor order by rand() limit 10;
create table sakila_bck.actor_film select * from sakila.actor_film where actor_id in (select actor_id from sakila_bck.actor);
-- Note: The create table xx select * from yy does not create keys for backup table. By the way, if you want to retrieve a random number of rows, you can try the PREPARED statement to provide the `limit clause` with a randomly generated number beforehand.
The whole process is definitely not a pushover as we have to keep table relationship in mind. But once the job is done, we can safely use mysqldump to dump sakila_bck at DB-level.