Nearly all enterprise digital transformation projects include moving software to DevOps processes and platforms. The effort, cost and time it takes to be successful with these transformations depend on a lot of factors, but one thing stands out: as indicated in my white paper Unified Enterprise DevOps Platform Why? How? And What?, “An enterprise with many variations of DevOps platforms and tools is at best chaotic, and at worst, is a gross impediment to overall organization agility – undermining the very purpose that DevOps is meant to serve.”
In my recent blog Why a Unified Enterprise DevOps Platform is Necessary, I indicated that “The aggregate set of platforms and tools is like a Frankenstein’s monster, with many incompatible parts resulting in an awkward looking product that is hard to control, replicate and evolve.”
Success with DevOps requires frictionless interplay between the people, processes and technologies that need to cooperate for continuous deliveries to happen effectively. The existence of different DevOps solutions creates organizational barriers to unification, because different groups become dependent on their local variations.
A Unified Enterprise DevOps Platform Accelerates Time-to-Value
Culture barriers, process problems and DevOps platform “wars” all are factors impacting the progress of digital transformations. A unified enterprise DevOps platform would accelerate time-to-value, and realize efficient, safe operation for digitally transforming enterprises and organizations.
The following are general requirements for a unified enterprise DevOps platform.
● A new, unified approach that supports end-to-end capabilities for continuous delivery of software, as a managed service, for multiple streams of pipelines.
● Unified metrics that show business value, costs of the value stream and timing for end-to-end continuous delivery flow across multiple pipelines.
● Unified capabilities including visibility, orchestration, integration, security, governance, traceability and management of flow for application value streams across multiple pipelines.
● Capabilities to unify continuous testing/QA, troubleshooting and debugging.
● A new, unified approach to simplify and accelerate toolchains that support new cloud-native and cloud-adapted applications, infrastructures, machine learning, commercial off-the-shelf and open source software.
A DevOps Platform Blueprint
In my SlideShare DevOps-as-a-Service Values, I described a four-layer blueprint for a unified enterprise DevOps platform that satisfies these general requirements. The main elements in the blueprint are:
● Stakeholders – There are three types of stakeholders: business, appdev users and DevOps admin teams.
● Service access layer – Consists of portals and APIs for business, appdev and DevOps administration users and access security.
● Service layer – consists of user services and DevOps admin services.
- User Services include things that appdev users need to perform their application development and testing work. This includes capabilities for no-code automation and orchestration, repos for code, artifacts and data, multi-CI/CD pipelines and sandboxes, test data, configuration, continuous testing/QA, service level objectives (SLOs) for business and appdev users and troubleshooting and debugging.
- DevOps admin services include things that a DevOps team needs to administer and manage the platform. This includes capabilities for platform administration, the ability to import pipelines into the unified platform, integration with enterprise service catalogs, integration with enterprise issues and change request systems, SLOs for DevOps administrators and capabilities to implement governance policies-as-code.
● Resources layer – includes resources that the services need to run including application stacks, infrastructure resources, tools, infrastructure code and topology data.
One of the most important benefits of a unified platform is the ability to unify SLOs for each class of stakeholders. The SLOs need to be chosen to best fit the requirements of each application.
Business value can be demonstrated by the following SLOs:
● Productivity metrics.
● ROI data tracking.
● Agility of services – Lead time for changes and frequency of releases.
● Reliability of services – Change failure rate; time to restore services.
● Governance compliance and security.
Appdev user value can be demonstrated by the following SLOs:
● Platform availability.
● Service response time.
● User experience scores.
● Change request response time.
DaaS team value can be demonstrated by the following SLOs:
● Scope of applications – number of apps supported, number of users per month.
● Number of requests served per month.
Without a unified enterprise DevOps platform, all these SLOs would have to be implemented multiple times, once for each type of DevOps platform.
To realize DevOps-as-a-Service for enterprises with multiple related pipelines, each pipeline needs to be configured to work within a unified model where common tools process each application and microservice continuous delivery task in parallel. Unified control and monitor points are needed – and should be exposed by the unified platform – to realize SLOs in a unified way. Without a unified platform, all of these SLOs will need to be implemented separately, for each platform, and then another system will be needed to aggregate them into a common view.
The SLOs will indicate where the biggest bottlenecks are for a particular pipeline. In many enterprise pipelines, the biggest bottleneck is testing, troubleshooting test failures and debugging to resolve the failures. It is a myth that continuous testing is just the execution of automated tests. The whole range of “testing things” includes test planning, test resource scheduling, test design, test implementation, test environment management, test execution, test results reporting, test result analysis, troubleshooting, debugging, fix retest and test management, and all can be sources of bottlenecks. These “testing things” are all sources of bottlenecks for every stage of a DevOps pipeline, for all Three Ways of DevOps. Without a Unified enterprise DevOps platform, all these testing things must be addressed separately, in accordance with the different pipeline platforms.
Testing, troubleshooting and debugging tasks are a major source of bottlenecks for most DevOps value streams. A unified platform improves continuous testing/QA, troubleshooting and debugging as follows:
● Continuous testing/QA – Orchestration of test environments, execution of tests and reporting test verdicts is unified.
● Troubleshooting – Failures are displayed in a consistent format, including trends correlated to pipeline and application log data, to simplify analysis of test failures.
● Debugging – Causes of failure can be isolated with intelligent analysis of troubleshooting results, and the platform can quickly install the fixes and orchestrate the test environment to verify the fix is working.
These testing, troubleshooting and debugging advantages have a major impact on the ability to accelerate digital transformations.
What This Means
Digital transformation of many enterprises and organizations are being seriously impeded by having too many DevOps platforms and tools. A unified enterprise DevOps platform resolves the technical problems of broken, leaky, “Frankenstein” toolchains and processes. Enlightened leaders that understand the importance and the broad impacts of unified enterprise DevOps platform are wise to recognize the sense of urgency surrounding a migration to a unified platform, and take action to stop the proliferation of too many platforms.
To realize a unified enterprise DevOps platform in an enterprise takes more than simply buying and installing a new platform and telling everyone to start using it in place of what they are used to. With multiple entrenched DevOps platforms in place, there are practical and political considerations that must be overcome. Introducing any changes across a large group of applications in an enterprise can be chaotic, and will often be more costly and take too much time if the changes are not well-coordinated in accordance with a carefully crafted strategy and implementation roadmap that leaders and team members are aligned with.