Many of us, old MySQL DBAs used Seconds_Behind_Source from SHOW REPLICA STATUS to find out the status and correct execution of (asynchronous) replication.

Please pay attention of the new terminology. I’m sure we’ve all used the old terminology.

However, MySQL replication has evolved a lot and the replication team has worked to include a lot of useful information about all the replication flavors available with MySQL.

For example, we’ve added parallel replication, group replication, … all that information is missing from the the good old SHOW REPLICA STATUS result.

There much better ways to monitoring and observing the replication process(es) using Performance_Schema.

Currently in Performance_Schema, there are 15 tables relating to replication instrumentation:

+------------------------------------------------------+
| …

In this new article about how to find the info when using MySQL Database Service on Oracle Cloud Infrastructure, we will learn about the query accelerator: HeatWave.

With HeatWave, you can boost the performance of your MySQL queries, providing your applications with faster, more reliable, and cost-effective access to data.

HeatWave is a high-performance in-memory query accelerator for MySQL Database Service on Oracle Cloud Infrastructure. It is designed to accelerate analytics workloads (OLAP) and increase the performance of your MySQL databases by orders of magnitude. This is achieved through the use of in-memory processing, advanced algorithms, and machine learning techniques to optimize query performance. If identified by the optimizer, OLTP requests can also be accelerated using HeatWave.

Today we will try to answer the following questions:

1. Can I use HeatWave ?
2. Is HeatWave enabled ?
…

If like me you are an old experienced MySQL DBA, to analyze your MySQL workload, you certainly have used the slow query log with long_query_time set to 0.

The slow query log is a file that contains all the queries whose execution time is greater than the value of long_query_time. This file can be huge and uses up all available disk space on busy systems.

Constantly writing to that file can also add an overhead to the server.

For those reasons, the slow query log is not available in MySQL Database Service (HeatWave) in OCI.

Plan B ?

As a DBA, what are my options for finding the queries that need to be optimized? As usual, the queries that consume the most time are the ones that need some attention

It can be a very long query or a short query executed too many times.

Currently the MySQL DBAs use …

Recently, I saw many interest in understanding and getting information about database locking.

MySQL InnoDB’s locking can be complex and having an overview not always simple.

For more information about how InnoDB locking is working, I can only recommend this excellent series of articles by Kuba:

• https://dev.mysql.com/blog-archive/innodb-data-locking-part-1-introduction/
• https://dev.mysql.com/blog-archive/innodb-data-locking-part-2-locks/
• https://dev.mysql.com/blog-archive/innodb-data-locking-part-2-5-locks-deeper-dive/
• …

Last Tuesday, it was the very first session of DB AMA, Morgan Tocker made a nice presentation of MySQL Performance_Schema and illustrated it with some nice queries to get Meta et Data Locks.

As those queries were not that simple to write or at least to remember, I thought it might be a good idea to add them to MySQL Shell, the best MySQL DBA Tool !

I’ve then added a new method to the check plugin: getLocks().

Let’t see it in action:

As you can see, this is a small extension that can offers you a better view of what’s locked per transaction.

You can find several MySQL Shell Extension directly on github: …

Today, we will cover a totally different MySQL Shell plugin: InnoDB.

Currently only 3 methods have been created:

Those related to the Table space fragmentation, have already been covered in this recent article.

Let’s discover the getAlterProgress()method. This method allows us to have an overview of the progress of some alter statements status like:

• stage/innodb/alter table (end)
• stage/innodb/alter table (flush)
• stage/innodb/alter table (insert)
• stage/innodb/alter table (log apply index)
• stage/innodb/alter table (log apply table)
• stage/innodb/alter table (merge sort)
• stage/innodb/alter table (read PK and internal sort)
• stage/innodb/alter tablespace (encryption)

This is an output of the method:

As …

In the first part of this article related to the check plugin, we discovered information retrieved from the binary logs. This part, is about what Performance_Schema and SYS can provide us about the queries hitting the MySQL database.

Currently, 3 methods are available:

• getSlowerQuery()
• getQueryTempDisk()
• getFullTableScanQuery()

The method’s name should be self explaining.

This is an overview of the parameters for each methods:

ext.check.getSlowQuery()ext.check.getQueryTempDisk()ext.check.getFullTableScanQuery()

Some methods allow a select parameter if only SELECT statements should be returned.

When only one query is returned (default), it’s also possible to …

Answering this question is not easy. Like always, the best response is “it depends” !

But let’s try to give you all the necessary info the provide the most accurate answer. Also, may be fixing one single query is not enough and looking for that specific statement will lead in finding multiple problematic statements.

The most consuming one

The first candidate to be fixed is the query that consumes most of the execution time (latency). To identify it, we will use the sys schema and join it with events_statements_summary_by_digest from performance_schemato retrieve a real example of the query (see this post for more details).

Let’s take a look at what sys schema has to offer us related to our mission:

> show tables like …

Recently, PeterZ pointed a huge difference in memory usage of MySQL 8.0 compare to MySQL 5.7. This can be an issue for small instances if the same configuration for buffers like the buffer pool are not changed.

As explained in Peter’s article, this can lead to the awakening of the so feared OOM Killer !

MorganT, pointed accurately in his comment what is the source of such difference and how this was then caused by the new instrumentation added in MySQL 8.0.

Nothing is free, even as a …

With MySQL Shell 8.0.16, a new very interesting feature was released: the Reporting Framework.

Jesper already blogged about it and I recommend you to read his articles if you are interested in writing your own report:

• https://mysql.wisborg.dk/2019/04/26/mysql-shell-8-0-16-built-in-reports/
• https://mysql.wisborg.dk/2019/04/27/mysql-shell-8-0-16-user-defined-reports/

I this post, I will show you one user-defined report that can be used to monitor your MySQL InnoDB Cluster / Group Replication.

Preparation

Before being able to use the report, you need to download 2 files. The first one is the …