[DZone Research] Microservices and DevOps Challenges

This article is part of the Key Research Findings from the  2018 DZone Guide to DevOps: Culture and Process .

Introduction

For this year's DZone Guide to DevOps, we surveyed 549 software professionals to find out their thoughts on various aspects of DevOps. In this article, we take a look at what respondents told us about microservices and the challenges they face.

Microservices

Microservices architectures have been a hot topic for years now, but adoption of microservices has been less frenetic. In 2016, 38% of respondents of our DevOps survey said their organization was using microservices in some capacity. In 2017’s survey results, we saw this increase to 44%. This year’s survey showed more than half (54%) of respondents using microservices — consistent with the results of the recent survey we conducted for our Guide to Microservices, where 53% of respondents said their organization was using microservices architectures for its applications.

Respondents saying their organization uses microservices were much more likely to believe their organization had achieved DevOps goals. Specifically, 74% of respondents using microservices architectures believed their organization had achieved Continuous Deliver, as opposed to 46% not using microservices; and 66% of microservices users said they believed their organization had achieved Continuous Integration, versus 41% of respondents not using microservices. Also, respondents using microservices architectures were much more likely (between 10-26%) than those not using microservices to say they broke their builds into stages (15% more likely); practiced performance and security issue detection (11% and 15%); automated performance testing (10%); conducted dependency, compliance, and code quality checks (13%, 12%, 26%); practiced code reviews (12%); and automated their checks to proceed with code commits and deployment (19%). 66% of respondents performing push-button deploys also said they were using microservices.

Challenges

A few CD pipeline pain points seemed to get less painful in the past year. Responses noting requirements issues as a pain point dropped 4% (23% to 19%), and challenges with regression testing (32% to 23%), performance testing (27% to 21%), and user acceptance testing (31% to 24%) all fell. The biggest CD pain points, however, did not see significant changes from last year. Environment configuration and setup is still the most common challenge, and only dropped 1% from last year (56% to 55%); the second-place pain point, coordination of team members and resources, also only fell 1% (34% to 33%).

Respondents believing their organizations have not achieved Continuous Delivery showed that barriers to entry for CD are shifting. While corporate culture – the most common barrier from 2017 – remained at the top of the list, it dropped 8% from last year’s results (53% to 45%). Support from management as a barrier also decreased 8% from last year (30% to 22%), and budget as a barrier fell 5% (27% to 22%). The barrier of integrating automation technologies into the SDLC, on the other hand, increased drastically from 2017, increasing from 26% to 40%.

Conclusion

Given that microservices is touted for its ability to quicken development time and ease many burdens of developing in a monolithic pattern, does the field of microservices hold the key to easing some of the pain points given above? 

Give us your thoughts in the comments?

This article is part of the Key Research Findings from the  2018 DZone Guide to DevOps: Culture and Process .