Over a million developers have joined DZone.
{{announcement.body}}
{{announcement.title}}

[DZone Research] Developing in DevOps

DZone 's Guide to

[DZone Research] Developing in DevOps

We explore what it's like to develop software in a DevOps-based working environment, and the pros and cons of this development approach.

· DevOps Zone ·
Free Resource

This article is part of the Key Research Findings from the 2019 DZone Guide to DevOps: Implementing Cultural Change.

DevOps and microservices are a natural pair, given their impetus on continuous, component-based development rather than a monolithic-based approach. As such, we have seen a growth in the percentage of respondents using a microservices-based architecture for their applications in production environments. In our 2018 survey, 29% of respondents claimed to use microservices in production; in this year’s survey, the adoption rate of microservices rose to 39%. We also saw a decrease in the number of respondents not using microservices. In 2018, 31% told us they were not using microservices but were considering them, and another 15% said they were not considering microservices at all. This year, these numbers fell to 27% and 13% respectively. Interestingly, when we compare this data to the two most popular types of developers reported in our survey (web app and business enterprise app), we find web app developers are more likely to use microservices. 42% of respondents developing web applications reported using microservices in production, whereas 38% of business enterprise developers use microservices in production. Despite this slight disparity in microservices adoption in production environments, web app (19%) and business app (18%) developers proved almost equally likely to adopt microservices in development. If you’d like to explore microservices more, check out the 2018 DZone Guide to Microservices.

In order to maintain and commit changes to code, respondents reported two main tools: GitHub and GitLab. Interestingly, despite the enormous popularity of GitHub in the software industry, the percentage of respondents using GitHub fell slightly from 37% in 2018 to 34% for this year’s survey. GitLab, however, saw a nice jump in adoption rates, rising from 23% in 2018 to 33% this year. When it comes to the tools used to deploy code, 64% report using the same deployment tool across all phases of the SDLC (development, QA, and production). This rose slightly from last year, when 61% of respondents reported using the same deployment tool across all phases of the SDLC.

No matter the tools used, almost half (47%) reported their main project’s build execution time to be less than 10 minutes. On a more granular level, 25% reported a build execution time of two to five minutes, and 22% reported a build execution time of six to 10 minutes. When we compare these numbers to our stats on web app and business enterprise app developers, we find that 28% of web application developers report a slightly faster build execution time. 28% of web app developers told us they have a build execution time of two to five minutes, whereas 23% of enterprise business app developers claimed a two- to five-minute build execution time.

When things go wrong, however, there seem to be two main causes. 75% of respondents told us that application errors/issues were the leading cause for rollbacks and/or hotfixes in their organization, and 52% reported environmental errors or issues. The mean number of rollbacks required among our general survey, however, proved rather low. Averaging out all the responses received, we find that 15% of deployments need rollbacks or hotfixes. And, continuing this positive trend, we find that the typical time to restore service when something goes wrong in production has gone down since last year. In 2018, 24% of respondents reported a time window of less than one hour, and this year, 28% of respondents reported the same time frame.

This article is part of the Key Research Findings from the 2019 DZone Guide to DevOps: Implementing Cultural Change.

Topics:
devops ,dzone research ,devops development ,devops developers ,devops pipeline

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}