Version Control (Concept)

Description

A guide to implementing the right version control practices for reproducibility and traceability.

Tags

capability, technical, core

Documentation

Version control systems like Git, Subversion, and Mercurial provide a logical means to organize files and coordinate their creation, controlled access, updating, and deletion across teams and organizations. Version control is closely related to automation. In fact, automation and continuous integration rely on these files for the source code of the automation itself, as well as the configuration to be automated and the data to be distributed.

In order to improve software delivery, teams need to use version control for source code, test and deployment scripts, infrastructure and application configuration information, and the many libraries and packages they depend upon. In the version control system, teams must be able to query the current (and historical) state of their environments. Version control also offers direct benefits such as disaster recovery and auditability.

Research shows that comprehensive use of version control, among other capabilities, predicts continuous delivery. In particular, version control helps you meet these critical requirements: * Reproducibility. Teams must be able to provision any environment in a fully automated fashion, and know that any new environment reproduced from the same configuration is identical. A prerequisite for achieving this goal is having the scripts and configuration information that are required to provision an environment stored in a shared, accessible system. * Traceability. Teams should be able to pick any environment and determine quickly and precisely the versions of every dependency used to create that environment. They should also be able to compare two versions of an environment and see what has changed between them.

These capabilities give teams several important benefits:

• Disaster recovery. When something goes wrong with an environment—for example, a hardware failure or a security breach—teams need to be able to reproduce that environment in a deterministic amount of time in order to be able to restore service.
• Auditability. To demonstrate the integrity of the delivery process, teams must be able to show the path backward from every deployment to the elements it came from, including their version. You enable this through comprehensive configuration management combined with deployment pipelines.
• Higher quality. The software delivery process is often subject to long delays waiting for development, testing, and production environments to be prepared. When this preparation can be done automatically from version control, teams can get feedback on the impact of their changes more rapidly, enabling teams to build quality into their software.
• Capacity management. When teams want to add more capacity to their environments, the ability to create reproductions of existing servers is essential. This capability enables the horizontal scaling of modern cloud-based distributed systems.
• Response to defects. When teams discover a critical defect, or a vulnerability in some component of their system, they need to release a new version of their software as quickly as possible. Storing all artifacts in version control means teams can roll back to a previously verified working state quickly and reliably.

As environments become more complex and heterogeneous, it’s progressively harder to achieve these goals. It’s impossible to achieve perfect reproducibility and traceability for a complex enterprise system (at a minimum, every real system has state). Thus, a key part of configuration management is working to simplify the architecture, environments, and processes to reduce the investment required to achieve the expected benefits.

Other Relations

From	Name	To	Description
Version Control	enables	Fast Flow

Navigation

List of views in namespace

List of all Views

(generated by Overarch with template docs/node.md.cmb)