Liquibase: example of automated roll changes on the relational database

Instead of a Preface


the Article will be interesting for those who have ever thought about the issue of rolling forward the changes (patch) to the relational database. The article will not be interesting to those who have already mastered and uses Liquibase. The main purpose of this article is the reference to the repository use case. As an example I chose rolling sample schema HR Oracle database (list of all supported databases) — anyone can download the repository and to play at home. The desire to demonstrate an example caused by the discussion at the resource sql.ru.


What Liquibase


What Liquibase can be found on official website of the product. I want to mention a couple of good articles on this resource:
Manage database migrations with Liquibase
Using Liquibase without a headache. 10 tips from my experience of developing real


Why I use Liquibase


I chose this instrument because:


1) the Tool keeps track of which changeset-s has already been applied to the instance DB and only those rolls that have not been rolled and which still need to concatit. If in the process of rolling forward the application of any change dropped by error, after the cause you restart the roll forward and Liquibase continues execution from that changeset and where they stopped.


2) Ability to set changeset-attributes runOnChange and runAlways greatly simplifies change management, in particular, recreatable objects.


3) the context Property allows you to execute/not to execute the changeset-s depending on the current environment (for example, to run unit tests on the prod).


It was not a full feature list.


Repository


He here. It provides "hard" (tables, indices, integrity constraints) and "soft" (triggers, procedures, views) objects changeset-s with the tags sql and sqlFile, attributes runOnChange and runAlways and without.


Which is not in the repository


Due to lack of need in the repository there is no such useful features/steps that I usually use in my projects:


the
    the
  • Preconditions — allows you to specify the execution condition changeset-a;

    • Compiling schema objects at the end of the roll. In Oracle's dbms_utility.compile_schema(user, false); the

  • Run unit tests.
Article based on information from habrahabr.ru

Комментарии

Отправить комментарий

Популярные сообщения из этого блога

A Bunch Of MODx Revolution + LiveStreet

When going to write your topic, puzzled where to write it? On the one hand part of the site written in MODx, on the other hand the other part written on LiveStreet, and on the third hand these are crutches that at the time of Abnormal programming to write. It turned out that the Habr was preparing for such an eventuality and made the ability to write to multiple blogs, separate respect! :-) Well, now actually to the basic question: At the time I wrote and launched coupon service itself is written in MODx Revolution (of course, because I am an ardent fan of this engine and whatever you read, always write on this framework). So, the project grew and developed a lot of functionality, it was written (the benefit of MODx allows to solve any problem), but that's just me zamisliti on the site and more social programs-blog a La ... run, yeah to all of our users were there, Yes to reviews stock there got, the reviews were in stock, and in the social sphere, and here and there t...
Далее...

Monitoring PostgreSQL with Zabbix

Изображение
PostgreSQL is a modern, dynamic DBMS with a very large set of possibilities which allow to solve a wide range of tasks. Using PostgreSQL as a rule refers to a very critical segment of the it infrastructure associated with the processing and data storage. Given the special place of the DBMS infrastructure and the criticality of tasks, the question of monitoring and appropriate control of the DBMS. In this respect, PostgreSQL has a wide internal means of collecting and storing statistics. Collected statistics allows to obtain quite a detailed picture about what is happening under the hood during operation of the DBMS. These statistics are stored in special system tables, views, and constantly updated. Performing regular SQL queries to these tables it is possible to obtain a variety of information about databases, tables, indexes and other DBMS subsystems. Below I describe the method and tools for monitoring PostgreSQL monitoring system Zabbix. I love this monitoring system becaus...
Далее...

PostgreSQL load testing using JMeter, Yandex.Tank and Overload

Изображение
a Few words to start so, this manual is suitable for those looking for a versatile Toolkit for testing large pool of systems and solving most problems on stress testing. This article is intended for beginners in this business, so I will try to Refine and simplify the process. Discuss briefly each of the elements in our bunch and move on to their primary installation and configuration: the Apache JMeter – load testing tool, which is able to carry out tests for JDBC connections, FTP, LDAP, SOAP, JMS, POP3, IMAP, HTTP and TCP out of the box and many other protocols and solutions using various plugins. the Yandex.Tank is a cloud – based tool for load testing uses different load generators, including and JMeter. the Yandex.OverLoad – the service for easy monitoring and analysis of servers under load. the Installation and configuration the key to success for proper installation of JMeter is the correct java installation. At this stage it is worth saying that all the manip...
Далее...