Steady State (stability pattern)
Steady State ( German steady state ) is a stability pattern in the information technology . The steady state pattern is used in particular for container virtualization.
functionality
Steady State means that a system in a production system must not be changed. Instead of updating the system, a new version of the system is provided.
Changes to the system are not permitted. As far as possible, log files are not stored on the production system, but on a central system that aggregates the data. This statelessness means that the system can be replaced at any time without loss of data.
Versioning
The old version of the system should also be kept ready in the form of a Named Stable Base or a Virtual Server Snapshot , so that in the event of an error in the new version of the system, you can switch back to the original version.
advantages
Using a steady state, it is possible to operate the old and the new version of the system at the same time. This also makes it possible to continue to support old clients and to verify the new version as part of an A / B test before switching to it. Furthermore, be zero downtime - migrations possible.
disadvantage
Since the systems are not updated automatically, any security gaps found in the system in the meantime can be exploited by hackers . It is therefore necessary to automate the provision of new versions of the infrastructure and services, as well as their functional verification by means of software tests , to the greatest possible extent in order to enable regular and rapid updates with patches .
swell
- ↑ a b c Michael T. Nygard: Release It! Design and Deploy Production-Ready Software. O'Reilly, 2007, ISBN 978-0-9787392-1-8 , 5.4 Steady State (English, 326 pages).