The latest from the MySQL community

So I tried to do my first set of benchmarks and testing on RocksDB today, but I ran into a problem and had to file a bug:
https://github.com/facebook/mysql-5.6/issues/365

MySQL @ Facebook RocksDB appears to store at least 2x the size of the volume of changes in a transaction. I don't know how much space for the row + overhead there is in each transcation, so I'm just going to say 2x the raw size of the data changed in the transaction, as approximation. I am not sure how this works for updates either, that is whether old/new row information is maintained. If old/row data is maintained, then a pure update workload you would need 4x the ram for the given transactional changes. My bulk load was 12GB of raw data, so it failed as I have only 12GB of RAM in my test system.

The workaround (as suggested in the bug) is to set two configuration …

In previous blog posts we explained how to perform certain actions in a MySQL database set up as a document store using Connector/Node.js. In this blog post we are going to use some of the examples covered to explain how to start working with an application created with Express.js, AngularJS, Node.js, and MySQL Connector/Node.js.

Required

• Node.js
• Express.js
• AngularJS
• MySQL Connector/Node.js
• MySQL Server 5.7.12 or higher

Optional

• …

InnoDBOnline buffer pool resize:

• Configure innodb_buffer_pool_size offline (at startup) or online, while the server is running 

• The operation is performed in chunks

• Chunk size is defined by the innodb_buffer_pool_chunk_size configuration option      

• Buffer pool size always be equal to or a multiple of innodb_buffer_pool_chunk_size * innodb_buffer_pool_instances

• Default innodb_buffer_pool_chunk_size is 128M

Reference:


https://dev.mysql.com/doc/refman/5.7/en/innodb-buffer-pool-resize.html



Online RENAME INDEX operation:

• ALTER TABLE RENAME INDEX requires only meta-data change.

Reference:

…

Prepare the RPM repository server

Install the software to create a RPM repository server (on an Ubuntu web server):

shell> sudo apt-get install createrepo

Create the directory structures and initialize the repositories:

shell> sudo mkdir -p /var/www/html/repos/centos/7/os/x86_64
shell> sudo createrepo /var/www/html/repos/centos/7/os/x86_64/

Copy the MySQL Enterprise RPM packages to the repository:

shell> sudo cd /var/www/html/repos/centos/7/os/x86_64/
shell> sudo unzip -v /download/V790254-01.zip   # MySQL Enterprise Server 5.7.16
Archive:  /download/V790254-01.zip
   Length  Method      Size Cmpr    Date    Time   CRC-32   Name
---------  ------  -------- ---- ---------- ----- --------  ----
 23979568  Stored  23979568   0% 2016-09-29 16:56 ae6693b1  mysql-commercial-client-5.7.16-1.1.el7.x86_64.rpm
 45327096  Stored  45327096   0% 2016-09-29 16:56 bb83c965 …

The pt-online-schema-change tool has been a workhorse for years, allowing MySQL DBAs to alter tables with minimal impact to application workload, and before MySQL had native online alter capabilities. Although consistently grateful for the tool’s existence, I never liked the messiness of having to use and clean up triggers, and most DBAs have a horror story or two to tell about using any of the online alter methods.

When Github’s online schema transmogrifer (gh-ost; https://github.com/github/gh-ost) was released, then, I jumped on the chance to test it out.

Following are my testing notes. I tried out the flags available to run gh-ost from different locations in a replication setup, both functionally and under load. The load applied was via mysqlslap, using a …

MySQL Cluster uses a distributed architecture where we have one primary
replica and one or more backup replicas of all data in the cluster.
In previous versions we have not supported reading from the backup
replicas. The reason for this comes from the transaction algorithm in
the NDB storage engine. When the transaction have committed we have
unlocked the rows in the primary replica, but the locks in the backup
replica is unlocked in the complete phase.

The impact of this is that we could potentially perform an update
of a row and then immediately read it and not see our own write. To
avoid this problem we have avoided to read the backup replica.

If it is important to read the backup replica then we need to ensure
that transactions are not reported back to the user until the complete
phase.

When we decided to support read from …

This blog post discusses the ramifications of STRICT mode in MySQL 5.7.

In short

By default, MySQL 5.7 is much “stricter” than older versions of MySQL. That can make your application fail. To temporarily fix this, change the

SQL_MODE

 to

NO_ENGINE_SUBSTITUTION

 (same as in MySQL 5.6):

mysql> set global SQL_MODE="NO_ENGINE_SUBSTITUTION";

MySQL 5.7, dates and default values

The default

SQL_MODE

 in MySQL 5.7 is:

ONLY_FULL_GROUP_BY,STRICT_TRANS_TABLES,NO_ZERO_IN_DATE,NO_ZERO_DATE,ERROR_FOR_DIVISION_BY_ZERO,NO_AUTO_CREATE_USER,NO_ENGINE_SUBSTITUTION

That makes MySQL operate in “strict” mode for transactional tables.

“Strict mode …

The Theory

Generated Columns is a feature released on MySQL 5.7. They can be used during CREATE TABLE or ALTER TABLE statements. It is a way of storing data without actually sending it through the INSERT or UPDATE clauses in SQL. The database resolves what the data will be.

There are two types of Generated Columns: Virtual and Stored. They work with:

• mathematical expressions (product_price * quantity)
• built-in functions (RIGHT(), CONCAT(), FROM_UNIXTIME(), JSON_EXTRACT())
• literals (“2”, “new”, 0)

Besides that, they can be indexed but they don’t allow …

In this blog post, we’ll explain how to update the signing key for Percona Debian and Ubuntu packages.

Some of the users might have already noticed following warning on Ubuntu 16.04 (Xenial Xerus):

W: http://repo.percona.com/apt/dists/xenial/InRelease: Signature by key 430BDF5C56E7C94E848EE60C1C4CBDCDCD2EFD2A uses weak digest algorithm (SHA1)

when running apt-get update.

Percona .deb packages are signed with a key that uses an algorithm now considered weak. Starting with the next release, Debian and Ubuntu packages are signed with a new key that uses the much stronger SHA-512 algorithm. All future package release will also contain the new algorithm.

You’ll need to do one of the …

This blog discusses the gradual end of MyISAM in MySQL.

The story that started 20 years ago is coming to its end. I’m talking about the old MyISAM storage engine that was the only storage provided by MySQL in 1995, and was available in MySQL for 20+ years. Actually, part of my job as a MySQL consultant for 10+ years was to discover MyISAM tables and advise customers how to convert those to InnoDB.

(Check your MySQL installation, you may still have MyISAM tables).

MySQL 5.7 still used MyISAM storage for the system tables in the MySQL schema.

In MySQL 8.0 (DMR version as of writing), the MyISAM storage engine is still available. But in a very limited scope:

• After …