Which DBMS to choose for a high-load project?

K

Khurshed Abdujalil2018-03-02 08:24:23

Database

Khurshed Abdujalil, 2018-03-02 08:24:23

Now there is a project in which there are millions of records. a relational database is used in mysql (version 5.5)
products (15 + million records)
category (1k)
tag (100k)
products_to_category (60 + million)
products_to_tag (60 + million)

but something slows down hard, now we are thinking of switching to another DBMS or upgrade to version 5.7, What to choose in order to at least somehow reduce the load?

ps slow down even more when sorting is enabled by creation date, if you sort it, the query will run slower by 5-10 seconds

Reply

Answer the question

In order to leave comments, you need to log in

3 answer(s)

E

Evgeny Kalibrov, 2018-03-02
@rework

For MySQL, the number of records you specify in the tables is not critical. Look for bottlenecks in queries and indexes. Try to debug your request through EXPLAIN, read more about it here - https://habrahabr.ru/post/211022/

V

Vasily Nazarov, 2018-03-02
@vnaz

15 million for MySQL with a competent scheme and queries is nothing at all.
However
1. It would be nice to have more memory (and this does not depend on MySQL or something more fashionable), and this memory must also be correctly "given away" (configured configs) to MySQL.
2. See specific SQL query.
Perhaps it makes sense to break it into several, without JOINs (JOINs often want a lot of memory, and if it is not enough, they slow down terribly). Especially when using sorting. (Your "when you turn on sorting" indirectly indicates that there is a problem here).

B

beduin01, 2018-03-02
@beduin01

Try to upgrade to MariaDB/Percona and tune the database.