The Problem

State Snapshot Transfer can be a very long and expensive process, depending on the size of your Percona XtraDB Cluster (PXC)/Galera cluster, as well as network and disk bandwidth. There are situations where it is needed though, like after long enough node separation, where the gcache on other members was too small to keep all the needed transactions.

Let’s see how we can avoid SST, yet recover fast and without even the need for doing a full backup from another node.

Below, I will present a simple scenario, where one of the cluster nodes was having a broken network for long enough that it will make …

Problem Statement

Having backups of binary logs is fairly normal these days. The more recent binary logs are copied offsite, the better RPO (Recovery Point Objective) can be achieved. I was asked multiple times recently if something could be done to “stream” the binary logs to S3 as close to real-time as possible. Unfortunately, there is no readily available solution that would do that. Here, I show what can be done and also show a proof of concept implementation, which is not suitable for production use.

In this example, the instance has two binary log files (mysql-bin.000001 and mysql-bin.000002) already closed and mysql-bin.000003 being written. A trivial solution for backing up these binary log files would be to back up just the closed ones (the one that is not written). The default size of the binary log file is 1 GB. This means with this solution we would have a 1 GB binlog not backed up in the worst-case scenario. On …

As some of you likely know, I have a favorable view of ZFS and especially of MySQL on ZFS. As I published a few years ago, the argument for ZFS was less about performance than its useful features like data compression and snapshots. At the time, ZFS was significantly slower than xfs and ext4 except when the L2ARC was used.

Since then, however, ZFS on Linux has progressed a lot and I also learned how to better tune it. Also, I found out the sysbench benchmark I used at the time was not a fair choice since the dataset it generates compresses much less than a realistic one. For all these reasons, I believe that it is time to revisit the performance aspect of MySQL on ZFS.

ZFS Evolution

In 2018, I reported ZFS performance results based on version 0.6.5.6, the default version available in Ubuntu Xenial. The present post is using …

Yesterday my friend and ex-colleague Sveta published a blog post on how to migrate ownership of stored routines, views and triggers in MySQL.

I usually agree a lot with her and even if I agree with what she wrote, there is one single point I would like to comment: use MySQL Shell !

In the blog post, Sveta recommends the use of mysqldump which I don’t. For any logical dump & load operations, I really recommend the use of MySQL Shell’s Dump & Load Utility ! It’s much faster, it has many nice options like hidden Primary Key creation, and is compatible with OCI.

And of course, MySQL Shell also provides a …

Benchmarking is a critical technique for delivering high performance software. The basic idea behind benchmarking is measuring and comparing the performance of one software version against another. Over the years, many benchmarking techniques have emerged, but we can broadly separate them into two categories: micro and macro benchmarks. Micro-benchmarks measure a small part of the codebase, usually by isolating a single function call and calling it repeatedly, whereas macro-benchmarks measure the performance of the codebase as a whole and run in an environment similar to what end-users experience.

“It would be nice to have an option, that would allow to suppress the DEFINER statement in the CREATE VIEW statements generated by mysqldump. This would help when transferring data structures between databases with different security models.” 

TLDR;

Use mysqlpump with option

--skip-definer

instead of

mysqldump

. The Story

This was requested as MySQL Bug #24680 on Nov 29, 2006. This feature request got large Community support. Even if we cannot see the number of people who voted for this request, the number of comments is impressive.

The request is very reasonable:

mysqldump

is widely used during application development and it is a very common practice to migrate database structure …

In the previous posts of this series, I presented how the Percona Support team approaches the analysis and troubleshooting of a MySQL server using a tried-and-tested method supported by specific tools found in the Percona Toolkit:

Inspecting MySQL Servers Part 1: The Percona Support Way

Inspecting MySQL Servers Part 2: Knowing the Server

Inspecting MySQL Servers Part 3: What MySQL?

Inspecting MySQL Servers Part 4: An Engine in …

Recently I blogged about how to easily deploy a LAMP application to MDS.

Using the Terraform module (OCI Resource Manager’s Stack) you also have the possibility to choose the PHP version to install:

But as you should already know, not all versions support the latest default authentication method in MySQL 8.0: caching_sha2_password [1], [2], [3].

Most of the PHP applications are using mysqli and mysqlnd …

Recently I blogged about how to easily deploy a LAMP application to MDS.

Using the Terraform module (OCI Resource Manager’s Stack) you also have the possibility to choose the PHP version to install:

But as you should already know, not all versions support the latest default authentication method in MySQL 8.0: caching_sha2_password [1], [2], [3].

Most of the PHP applications are using mysqli and …

The new MyDumper 0.10.7 version, which includes many new features and bug fixes, is now available.  You can download the code from here.

For this release, we have added several features like WHERE support that is required for partial backups. We also added CHECKSUM for tables which help to speed up the restore of large tables to take advantage of fast index creation, and more.

New Features:

• Adding metadata file per table that contains the number of rows #353
• Adding –where support #347 …