Recently we got a question about how one can perform a minor version upgrade of a Galera Cluster, and today we just want to say that it is reasonably straightforward. While upgrades are a feature of Galera Manager that we are most excited to see come soon, today, doing it manually on the command line is quite simple! In this post, we will go through upgrading a 9-node geographically spread cluster, from MySQL 8.0.23 to 8.0.26. We also have extensive documentation: Upgrading Galera Cluster.

We have 3 nodes each in Singapore, London and San Francisco, making the cluster comprise of a total of 9 nodes, as we can confirm by executing: show status like 'wsrep_cluster_size';.

Since we installed all this via the RPM package manager on CentOS 7 by adding the YUM repository to the …

In 2009, I showed an example of how to use the MySQL LOAD DATA INFILE command. Last year, I updated the details to reset the secure_file-priv privilege to use the LOAD DATA INFILE command, but you can avoid that approach with a simple Python 3 program like the one in this example. You also can use MySQL Shell’s new parallel table import feature, introduced in 8.0.17, as noted in a comment on this blog post.

The example requires creating an avenger table, avenger.csv file, a readWriteData.py Python script, run the readWriteData.py Python script, and a query that validates the insertion of the avenger.csv file’s data into the avenger table. The complete code in five steps using the sakila demonstration database:

• Creating the …

Working through our chapter on MySQL views, I wrote the query two ways to introduce the idea of SQL tuning. That’s one of the final topics before introducing JSON types.

I gave the students this query based on the Sakila sample database after explaining how to use the EXPLAIN syntax. The query only uses only inner joins, which are generally faster and more efficient than subqueries as a rule of thumb than correlated subqueries.

SELECT   ctry.country AS country_name
,        SUM(p.amount) AS tot_payments
FROM     city c INNER JOIN address a
ON       c.city_id = a.city_id INNER JOIN customer cus
ON       a.address_id = cus.address_id INNER JOIN payment p
ON       cus.customer_id = p.customer_id INNER JOIN country ctry
ON       c.country_id = ctry.country_id
GROUP BY ctry.country;

It generated the following tabular explain plan output:

 …

In reply to a question about how to conditionally drop an index on a table in MySQL. It appears the syntax doesn’t exist. However, maybe it does and I missed it. If I did miss it, I’m sure somebody will let me know. However, I simply have a dropIndexIfExists stored procedure for this type of database maintenance.

Below is my dropIndexIfExists stored procedure:

-- Conditionally drop the procedure.
DROP PROCEDURE IF EXISTS dropIndexIfExists;

-- Change the default semicolon delimiter to write a PSM
-- (Persistent Stored Module) or stored procedure.
DELIMITER $$

-- Create the procedure.
CREATE PROCEDURE dropIndexIfExists
( pv_table_name  VARCHAR(64)
, pv_index_name  VARCHAR(64))
BEGIN

  /* Declare a local variable for the SQL statement. */
  DECLARE stmt VARCHAR(1024);

  /* Set a session variable with two parameter markers. */
  SET @SQL := CONCAT('ALTER TABLE ',pv_table_name,'DROP INDEX ',pv_index_name);

  /* Check if the constraint exists. …

I was working through some tutorials for my students and noticed that there was a change in how a WHERE clause must be written against the information_schema.table_constraints table. It might have been made in an earlier release, I actually hadn’t checked it since 2014 when I wrote this early post on capturing MySQL Foreign Keys.

You could use the following WHERE case insensitive clause:

WHERE    tc.constraint_type = 'foreign key'

Now, you must use a case sensitive WHERE clause:

WHERE    tc.constraint_type = 'FOREIGN KEY'

I’d love to know why but I can’t seem to find a note on the change. As always, I hope this helps those looking for an answer.

When I went over my example of using the WITH clause to solve how to use a series of literal values in data sets, some students got it right away and some didn’t. The original post showed how to solve a problem where one value in the data set is returned in the SELECT-list and two values are used as the minimum and maximum values with a BETWEEN operator. It used three approaches with literal values:

• A list of Python dictionaries that require you to filter the return set from the database through a range loop and if statement that mimics a SQL BETWEEN operator.
• A WITH clause that accepts the literals as bind variables to filter the query results inside the query.
• A table design that holds the literals values that an analyst might use for reporting.

It was the last example that required elaboration. I explained you might build a web form that uses a table, and the table could allow a …

As an example for my class on the usefulness of Common Table Expressions (CTEs), I created three examples with Python. They extend an exercise in Chapter 9 on subqueries from Learning SQL by Alan Beaulieu. All of the examples work with the sakila sample database.

These bullets describe the examples:

1. Uses local variables and a range for loop and if statement that uses the variables to evaluate and add an element to the derived table (or query result set) from MySQL.
2. Uses a CTE with substitution variables from the Python program, which eliminates the need to evaluate and add an element to the query result set because the query does that.
3. Uses a table to hold the variables necessary to evaluate and add the element to the query result set.

This is the first Python program:

# Import the library.
import sys
import …

An interesting outcome of teaching SQL is discovering what skills new users require. One that I continuously rediscover is how to build a test case for various elements of SQL. This is a small article on querying with filters in the WHERE clause.

There are several of the exercises in Alan Beaulieu’s Learning SQL, 3rd Edition that would benefit from example setup. For example, Chapter 4 provides a snapshot of the payment table but doesn’t provide any instructions.

You can create an exercise_4_2 table with the following SQL statement if you plan to change the data:

CREATE TABLE exercise_4_2 AS
SELECT payment_id
,      customer_id
,      amount
,      payment_date
FROM   payment
WHERE  payment_id BETWEEN 101 AND 120;

Alternatively, you can create an exercise_4_2 view with the following SQL statement if you plan to only query the data:

CREATE VIEW exercise_4_2 AS
SELECT payment_id
, …

Since we just released Galera Manager, it seemed like a good idea to do a deployment on Amazon Web Services (AWS) using Red Hat Enterprise Linux 8 and MySQL 8.

We start by launching an EC2 instance and choosing Red Hat Enterprise Linux 8 (HVM) and most importantly, we only choose the 64-bit (x86) architecture as we do not support ARM just yet. We choose a t2.micro instance for Galera Manager since this is just a test instance, but it is well worth knowing that 10GB which is the default EBS size may be a little small if you plan on using this for sometime.

…

Sometimes life is too surreal. Like when somebody says, “How do you get title case in an Oracle database?” That’s when you know three things about the individual, while suppressing laughter. They’re not very experienced with SQL, likely lazy, and don’t read the documentation.

I had a little fun with somebody today by taking them down a small rat-hole. “Oh, gosh … ” I said, “… let’s write a function for that.” Here’s the joke function, like:

CREATE OR REPLACE
FUNCTION title_case
( string VARCHAR2 ) RETURN VARCHAR2 IS
BEGIN
  /* Change upper case to title case. */
  RETURN UPPER(SUBSTR(string,1,1)) || LOWER(SUBSTR(string,2,LENGTH(string)));
END title_case;
/

Then, we tested it with a query from the pseudo dual table:

SELECT title_case('incredible') AS "Proper Name" FROM dual;

It returned:

Proper Name
----------
Incredible

Then, I said “Oh, that’s not his …