Percona live is over, and we finally get some feedback about our presentations.

Initially I had 3 presentations to give:

• Comparing High Available Solutions With Percona XtraDB Cluster and Percona Server With Group Replication.
• Comparing Hash Join solution, the good, the bad and the worse. The New version 2021.
• Boosting MySQL NDB cluster & MySQL Innodb Cluster with PrxySQL V2

But when I saw we were having a lot of great submissions, I decided that it would be better for the conference and for Percona to drop 2 of them and leave some free slots for others.

That has always been my line of conduct, I do not think we should have speakers with multiple talks unless exceptions.  I know in the past I had been presenting multiple times, but that is why I am stating this now, more and more.

Anyhow, the remaining talk was Comparing High Available Solutions With Percona …

After Oracle released MySQL version 8.0.24 on April 20, 2021, our engineering team got started right away with merging our enhancements to prepare the corresponding 8.0.24 version of Percona Server for MySQL.

However, Oracle released MySQL version 8.0.25 shortly afterward on May 11, 2021, to fix a critical bug that we also observed during our initial testing and reported back to them.

Therefore, we have decided to skip releasing Percona Server for MySQL 8.0.24 both as a standalone product and a distribution as well as the …

Recently, I wrote several articles about how to deploy popular Open Source applications on Oracle Cloud Infrastructure and MySQL Database Service.

Today we will see how you can deploy your own LAMP stack application using the same technique where L will stand for a compute instance (and why not the Ampere always free trier?), A stays Apache and will run in that compute instance. M stands for MySQL Database Service and P for PHP.

As usual we start by deploying a Stack by just clicking on the deploy button from GitHub:

The we are directly redirected to OCI’s dashboard and we need to accept the Terms of Use:

As soon …

Matomo is a Google Analytics alternative for your websites. If you follow my blog, you know how easy it’s to deploy popular Open Source web solutions like WordPress, Joomla!, Drupal, Moodle, Magento on OCI. If this is not yet your case, please check this page: deploy on OCI.

All these solutions are using MySQL Database Service to store their data.

I’ve recently added a new stack to also deploy Matomo. Of course this can be a standalone installation to collect all your analytics from self-hosted websites, but today I will describe you how to use it with an existing stack we already deployed on OCI.

For this example, I deployed WordPress using the following stack: …

In this blog, we will discuss, how to setup MySQL NDB Cluster replication through backup and restore method. This is bit tricky but interesting. The scenario here is, say user have a standalone cluster up and running, later there is a need to have a replication setup with an another empty cluster without shutting down the existing running cluster i.e. zero downtime. Then this backup and restore method will come in handy.

Let’s create two MySQL NDB Cluster with the following environment, Here, one will be termed as ‘source’ cluster and the other one will be termed as ‘replica’ cluster.

• MySQL NDB Cluster version (Latest GA version)
• 1 Management Node
• 4 Data Nodes
• 1 MySQLDs
• Configuration slots for up to 4 additional API nodes

Replication schema diagram:
CLUSTER 'A'        …

Yesterday, Oracle announced a new range of Arm compute instances based on Ampere’s ARM processors

This is a very nice announcement ![1], [2] and something also very cool is that you are eligible for an Always Free instance up to 4 OCPUs, 24GB of RAM and 4Gbps network bandwidth !!

This is the max you can have and you define it during the shape selection:

I have already updated all the Stacks I propose to …

Computers are dumb. And they will do exactly what you ask them to do.  The trick often is to think as dumb as the computer. Sadly it is all to easy to assume that the computer is 'thinking' like you are and blunder into a head scratching puzzle.  Recently there was a post on MySQL Community Space Groundbreakers Developer Community site that shows that sometimes what is intended is not what you want but you are getting exactly what you asked. 

Quiz -- What happens if you run the following query?

SELECT concat('CREATE TABLE if does not exists sakila1.', 
    TABLE_NAME, 
    ' like sakila.', 
    TABLE_NAME,  ';') 
FROM information_schema.`TABLES` 
WHERE TABLE_SCHEMA = 'sakila'

A) You will …

In this blog, we will discuss about bidirectional replication between two NDB clusters. With this topology, user can do transaction simultaneously from both the clusters on the same database objects.

Let’s create two MySQL NDB Cluster with the following environment, Here, both will be termed as ‘source’ cluster as well as ‘replica’ cluster as we will do replication from both the sides.

• MySQL NDB Cluster version (Latest GA version)
• 1 Management node
• 4 Data nodes
• 2 MySQLDs
• Configuration slots for up to 4 additional API nodes

Replication schema diagram:

                    |-----------S1---------->>----------S3--------------|                   …

 In the 1990s I spent a few years studying requirements on databases used in 3G telecom networks. The main requirement was centered around three keywords, Latency, Throughput and Availability. In this blog post I will focus on Availability.

If a telecom database is down it means that no phone calls can be made, internet connections will not work and your app on your smartphone will cease to work. So more or less impacting each and everyone's life immediately.

The same requirements on databases now also start to appear in AI applications such as online Fraud detection, self-driving cars, smartphone apps.

Availability is measured in percent and for telecom databases the requirement is to reach 99.9999% availability. One often calls this Class 6 availability where 6 is the number of nines in the availability percentage.

Almost every database …

I found myself explaining the nuances of INSERT statements and whether you should use named or positional notation. While the class was on Zoom, I could imagine the blank stares in the silence of my headphones. Then, I had to remind them about mandatory (NOT NULL constrained) and optional (nullable) columns in tables and how an INSERT statement requires an explicit NULL value for optional columns when the INSERT statement isn’t inserting a value into that column.

Then, I asked if somebody could qualify the different types of INSERT statements; and what would happen if a table with a first_name and last_name column order evolves when a new DBA decides to restructure the table and uses a last_name and first_name column order in the new table structure. Only a couple of the students recalled using a column-list …