Automating IT infrastructure provisioning and other aspects of application development isn’t new to anyone. While many companies understand the importance, many are immature with respect to their approach to standardizing the orchestration and automation of the build, deploy, and release lifecycle. DevOps and Application Release Automation (ARA) are driving significant focus on improving collaboration and automation across IT, enabling businesses to address their need for speed, quality, and better cost dynamics.
Flexagon’s FlexDeploy is a DevOps and ARA product which standardizes, orchestrates, and automates the lifecycle for provisioning, building, deploying, and releasing applications, middleware, database changes, and other related artifacts across environments. Please see the FlexDeploy home page for additional information, www.flexagon.com/flexdeploy
While FlexDeploy helps achieve the benefits described, this is written to be vendor/product neutral, and generalizes the ROI and value associated with DevOps and ARA.
How does an automation framework/tool bring value and ultimately a return on investment? What are the key benefits that can help justify the focus?
This post helps you answer such questions, and in many cases, quantify the value in hard dollars. Flexagon views Application Release Automation spanning Provisioning, and the entire lifecycle of Build, Deploy, Test and Release of applications and related artifacts across non-production and production environments. To simplify, we will use the terms Deploy or Deployment to cover the entire Build, Deploy, and Release lifecycle.
Benefits can be broadly categorized across 6 areas:
- Reduce manual labor and time spent creating and maintaining deployment scripts and tools.
- Reduce time spent debugging and fixing deployment related errors.
- Reduce outages across environments, impacting both the productivity of your development and test teams and application disruption when a deployment fails in production.
- Increase speed of deployments and ability to deliver software to the market faster.
- Decrease time spent demonstrating compliance, or the impact of failing an audit.
- Improve employee satisfaction and retention by enabling Dev and Ops staff to shift their focus to higher value activities.
Not everyone prioritizes benefits the same, so you will need to decide for yourself what really matters for your organization. For example, a deployment related outage in production can have significant cost implications for some companies, and that company would therefore rank quality improvements and a reduction in deployment errors as a top priority. Others need to improve the speed of software delivery and business agility and subsequently focus on increasing the rate at which deployments can occur.
Reduce manual labor and scripting
Application Release Automation tooling eliminates manual work and eliminates or greatly reduces scripts via pre-built plugins. This is the easiest category of cost savings to convert into hard dollars. Some companies have people dedicated to handling the deployment and release process, while other spread this work across many individuals in the development and operations teams. In either case, you can calculate the hours spent performing manual work and also the time spent creating and maintaining the scripts and other tools that support the deployment processes.
Whether using manual and/or scripted processes today, make sure you consider the time required to update the processes and scripts over time. In many cases, the products you are using change their deployment techniques over time. Your current manual processes and other scripts will need to be updated accordingly. When using an Application Release Automation product, the burden of creating and maintaining the plugins/integrations is taken on by the vendor and not your internal staff.
The number and complexity of environments, technologies, projects, and applications can significantly increase the overall cost. When calculating the cost associated with your process and tools, make sure to capture all deployment related activity. For example, companies using Oracle Fusion Middleware might have WebLogic, SOA, E-Business Suite, and ODI products used across 2-6 non-production and production environments. Calculate the cost for each product, across all environments, and for all projects executed in a given year.
Reduce time spent debugging and fixing deployment related errors.
Manual and scripted processes are often error-prone and result in development, operations, and other IT support personnel spending significant time troubleshooting failures. Capture the details of failures across all environments and the time spent on error resolution activities.
While capturing the hours spent debugging and troubleshooting deployment related issues, it’s important to include times when the failures cause war-room like conditions. This includes the management team being involved until the problem is either rectified or the changes are rolled back to a prior version. The waste associated with those situations can cut across more than the technical staff directly involved with issue resolution, so make sure to capture those details.
Reduce deployment related outages across environments
Outages can impact productivity of development and test teams and have a direct business impact for both internal and external customers. The prior category captured the time spent getting the deployment related issues resolved by the technical team and other leaders overseeing the troubleshooting. This ‘environment outage’ category captures the cost related to the users impacted because an environment is not available. For example, when a deployment fails in a Test or QA environment, it might take several hours to debug and resolve the problem. Developers take a productivity hit because they can’t test their changes or develop new features. In many cases, testers (IT and business) are lined up to perform their test activities and are left in an unproductive state when the environment isn’t available.
When the outage occurs in production, the impact can be significant. For example, when a critical revenue generating ecommerce system isn’t able to accept customer orders, or the impact to customer satisfaction due to a customer support system being down. It is not always easy to quantify, however the negative impacts and costs are real and can be very hard to recover.
Increase speed of software delivery
As many companies put more focus on providing innovative products and services to their customers, the demand for agile processes for software delivery goes up dramatically. How can improvements in speed of delivery be quantified? You can compare the current deployment processes to automated processes, analyzing factors such as:
- Faster and higher quality deployments reduce development/test cycles. The hours/days saved as part of a project can be significant and result in software delivered to the market earlier. The results can have direct impacts to revenue generation, business operation efficiency, and customer satisfaction. Depending on your business and type of software being delivered, you can calculate the financial impact of delivering software to the market faster.
- Outage windows for deployments/releases – due to the automation, the time it takes to make a change will decrease the system downtime. This can be valuable, especially for revenue generating systems.
- Amount of change in a Release can increase because the deployments can be parallelized.
- If non-deployment related issues surface after a deployment/release, the rollback can occur faster, mitigating the impact to the end users.
Decrease time spent proving compliance and the impact of failing an audit
Standardizing and automating the deployment process results in visibility to the “who, what, when, and where” across all environments. The IT team is better positioned to demonstrate compliance to a controlled process, and improves the likelihood of passing an audit.
Improve employee satisfaction and retention
Deployment activities into production environments are often performed on nights and weekends. Lengthy and error-prone processes can result in challenging situations for key IT resources expected to support the deployments. The late-night activities can be mentally draining and increase risk, as tired and stressed developers and operations team members don’t perform consistently well under those conditions.
Enabling teams with processes and tools which improve the deployment lifecycle can have multiple benefits in terms of employee satisfaction and retention. First, it will increase the speed, quality, and efficiency of off-hours deployments, reducing employee stress. In addition, faster and higher-quality deployments can result in the ability to move changes from nights and weekends into normal business hours.
Secondly, automating the deployment process allows staff members to spend their time performing higher value and more enjoyable activities, therefore improving job satisfaction and retention. Increased retention has an additional impact on the company, as the cost of recruiting and training a new employee is high.
Manual and scripted processes are slow, error-prone, and cost a lot of time and money to build and maintain. With automation, teams are able to build and deliver higher quality software at lower cost and with less stress. It can be challenging to move away from something that is “good enough,” but making the move can pay off in spades! If you are ever interested in discussing the possibilities, don’t hesitate to reach out to me at [email protected]!