What ‘Good’ looks like
Designing an ideal landscape
You can’t hit a target you can’t see - for us, defining the desired target state of a project and its business outcome is just as important as the technical delivery. Our goal is always to define and reach ‘What Good Looks Like’ within each customer engagement, such as;
Zero-touch continuous deployments: Releases are pushed to production environments the moment the code change has passed it’s automated checks, security scans and QA tests without pressing a button, in an instant.
Security-by-Design baked into pipelines: Code integrity checks at the point of integration into the codebase, security scanning for vulnerabilities, secrets management and static code analysis are all methods that, when implemented, create tougher, more resilient products.
Automatic environment creation & destruction: Environments cost you every second they exist; therefore the ideal is CI/CD actions that prop up dynamic test environments from code, that immediately dissolve once its function is completed and the deployment progresses.
Continual monitoring and graphed dashboards: Extensive monitoring, reporting and alerting tools are in place, with key transactions and performance metrics well-defined and understood, feeding information back across all teams transparently following each release.
Fully automated testing and validation pipes: Individual component testing, regression testing, performance testing system integration testing and acceptance testing are fully scripted and automated to allow a clear runway between code-change and live build.
Microservices architecture without dependencies: Taking a fully modular approach to architectural design means each component exists and operates in isolation, communicating via a messaging system, dependant on nothing and easily identifiable as a point-of-failure when errors occur.
100% config managed Infrastructure-as-Code: All components standardised into version-controlled files and integrated into the build-test-release pipeline, automatic scalability as needed, self-rebuilding infra during fail states, new customers onboarded by filling out a short web form and pressing a button.
Breaking down your current state
Fast-moving start-ups without a clear long-term vision accrue technical debt. Being ‘Agile’ is no longer enough. If any of the following ring true within your business - and we like to think we’ve seen it all by now - you may be heading down the wrong path;
Manual releases with many steps: Documented SDLC involving many teams and responsible parties with numerous manual steps and branching. New builds splutter along slow, congested pipelines with no regular (or measured) release velocity, leading to waterfalled releases.
Security checks as a manual process: QA Team owns security testing as part of their other manual tests and checks in the Test phase. Manual security efforts are slow, prone to human error and require specialised skills that can present blockers in instances of staff sickness, vacation or other absence.
Virtual test & staging environments: Environments are manually cleaned, upgraded and maintained by a nominated environment manager. Images reverted back after each test cycle, external management tools (or spreadsheets) used to keep track of status and changes.
Manual reporting and siloed data-sharing: Performance metrics are pulled by nominated staff on request, limited visibility to teams. Logfiles pulled for support teams on request in the event of an incident. Reactive rather than proactive.
Staging and testing by QA team: Individual component testing, regression testing, performance testing system integration testing and acceptance testing are performed by team members using a checklist. Checks are thorough but slow, reducing release velocity and presenting blockers.
Monolithic, dependent architecture: Components are not decoupled from one another and failover on one will trigger errors in another. Large-scale downtime, product outages, critical failures occur as a result. Scalability requires code changes to implement, meaning product needs to be taken offline to upgrade.
Traditionally-managed infrastructure: Infrastructure components are built, maintained and monitored by conventional IT teams; servers, storage and other network resources must be provisioned on-request. Standardisation is a target but not a constant, lack of versioning makes traceability difficult.
Releaseworks® helps you find your company’s ‘Good’
Our consultative approach towards DevOps is highly structured and repeatable; fully chart out your current state, detail your desired target state, plot out the milestones necessary to get you there, and assign a timeline to it - your own personal journey to ‘Good’. We dream it, define it and deliver it.
Supercharge your DevOps capabilities with Releaseworks®
We design, build and support
Releaseworks® is the trusted partner for Cloud and DevOps.
Our service offering includes DevOps and Cloud strategy, Cloud Engineering, and 24/7 Global DevOps support.
We understand your needs
At Releaseworks®, we specialise in Cloud and DevOps only.
We work with start-ups and enterprise organisations, and are well-versed in the unique challenges of both environments.
Experts available on demand
Our team consists of permanent full-time employees based in the United Kingdom.
All Releaseworks® engineers hold one or more industry certifications, and are background checked according to the UK Government Baseline Personnel Security Standard (BPSS).