Quick hypothetical scenario: meet Dan, an application development executive at a large retailer. One morning, Dan was swamped with complaints about the new enterprise messaging system not working when he walked into work on what he thought would be a lovely day. He had sent out a note to the employee community announcing its availability the evening before. And Dan had spent the entire day testing the application and clearing it for a launch.
Dan was not surprised. Despite all checks, there could always be an undiscovered dependency that was not considered. In this case, it turned out to be a certain security policy implemented long ago to take care of a threat that no longer existed. It took Dan the better part of the week to find the cause and fix it.
That’s configuration drift.
Configuration drifts: Are they inevitable?
Every IT team spends considerable time ensuring that different environments in a software development lifecycle have the same configuration. Provisioning different environments for development, test, quality assurance and production takes weeks and involves coordination across teams.
Often enough there are manual changes in the development, test or quality assurance environment that are not conveyed to the production environment and the changes lead to errors. Each team may make slight adjustments to their environment that causes configuration drift, creating complexity and communication nightmares across teams. Much time, effort and expense is wasted trying to identify and fix subtle differences. But it doesn’t have to be this way.
A workable solution to this problem is to standardize a pattern for the full-stack hardware and software infrastructure—the complete application and middleware environment—and re-deploy that pattern repeatedly. This would mean each environment picks up the same pattern and the same configuration, thus avoiding configuration drift for any one or multiple environments. This also helps avoid unforeseen configuration issues for all development, test, quality assurance and production environments. And you can avoid the finger-pointing and wasted time trying to identify and fix issues at or after deployment.
Three forces to tame configuration drift
There are three forces to keep the configuration drift beast under control.
- Automation: Standardization and automation of provisioning and deploying app and middleware environment through patterns makes it easier to maintain consistent configurations across all environments.
- Synchronization: Patterns ensure that all application environments are in sync and error free throughout their lifecycle, thus streamlining dev, test, QA and production rollouts.
- Adoption of best practices: Configuration patterns set the standard application and middleware environment and reduce unpredictability.
A solution that can do all the above would be ideal.
IBM PureApplication: The forces unite to provide a single solution
IBM PureApplication is a set of offerings that converges compute, storage, networking, and a middleware and software stack—including PureApplication Software—into a preintegrated, preconfigured and pretested system. You can be up and running within hours, saving tremendous time, effort and resources versus purchasing, installing, configuring and coordinating patches across individual hardware and software components.
Common to every PureApplication offering is a set of best practices that are captured in patterns. Important configuration information is stored in a pattern, such as middleware deployments and connections to data sources. Pre-built patterns for popular enterprise workloads are available, and they can be easily customized for your unique app and middleware environment. This pattern can be executed with push-button ease to deploy the exact same environment to development, test, quality assurance and production environments. You can eliminate differences across environments and avoid future configuration drift.
Automated provisioning and configuration helps accelerate application delivery so you can get your app into production must faster. It also eliminates errors and reduce the time, effort and cost to identify and fix those errors caused by manual processes and configurations.
Check out the great return on investment of PureApplication from the latest Total Economic ImpactTM of IBM PureApplication from Forrester Research.
Share this post: