Pain Points and Risks for Developers
Software developers should be able to focus on building code rather than deploying it. But, Software deployment often can be a complex process that comes with pain points like: customer outages, less time spent on innovation, lack of consistency, and numerous manual steps. In short, software deployment remains time-consuming, risky, and unreliable.
Armory and Gartner Peer Insights sampled 200 engineering and operations professionals to better understand their current approach to software deployment, testing, and metric tracking.
Data collection: April 7 – June 16, 2022
Respondents: 200 engineering and operations professionals
How many different environments does your organization deploy to today?
For respondents who provided a number (n=156), the average number of environments was 21.8.
Manual steps and poor workflows are holding back many organizations from improving deployment frequency
Cited by 54% of respondents, the most common pain point as part of the deployment process is too many manual steps.
Which pain points do you currently experience as part of your deployment process?
Armory is designed to help you get from manual to fully automated safe, secure, and reliable deployments
Poor processes and workflows is the primary cause of current pain points in the deployment process for 31% of respondents.
What do you believe to be the primary cause of these pain points?
Development frequency is the top success metric for deployments, selected by over half (52%) of engineering and operations leaders.
What metrics do you track to measure
the success of your deployments?
Triggering Deployments from GitHub and GitLab and Leveraging Rolling Deployments Lead the Way
GitHub and GitLab is the most common way engineering and operations leaders trigger their deployments.
How do you trigger your deployments today?
Trigger Armory Continuous Deployment from any CI tool like GitHub Actions
Rolling deployments are used by 42% of respondents today, while canary deployments are the most popular deployment method leaders aspire to use in the future.
Which advanced deployment methods does your organization use or would you like to use?
Application health metrics monitored manually, while custom tests before deployment are mostly run in a staging environment
Nearly half (46%) of organizations manually monitor application health metrics after deployment. Just 37% use automated canary analysis.
How does your organization leverage metrics to analyze the health of your application?
The majority of respondents (54%) run custom tests in a staging environment before deploying to production.
Do you run custom tests before deploying to production?
Tightly coupled services are moderately or extremely important to most
64% of respondents rate tightly coupled services that can deploy and rollback together as moderately or extremely important to their organization. Almost one-quarter (24%) rate traffic management in their Kubernetes clusters as slightly or not at all important to their organization.
Please rate the following based on how
important they are to your organization:
Developers can only modify their own apps
Tightly coupled services that can deploy and rollback together
Traffic management solutions to manage your Kubernetes clusters
Download the PDF
Commit. Deploy. Repeat.
Focus on writing great code, not deploying it!