Replay Solutions, which conducted the survey, directed survey respondents to the DevOps Wikipedia page for the definition, but its working definition is the merging of developers, quality assurance and operations into an organizational unit. The benefits of such a merging are improved communications between the teams, faster provisioning time with fewer errors, and increased use of programmatic automation from server, storage and network provisioning, all the way up to application deployment and management.
The respondents were primarily from North America, and the titles covered developers to IT operations and from administrators to C-level. The company size by number of employees was evenly split among organizations with more than 2,000 employees and those with less then 2,000 employees. Approximately 25 percent of respondents say they have dedicated positions titled DevOps. The survey respondents appear to identity more with development than with operations, which makes sense considering that Replay Solutions focuses on software quality assurance and resolution.
For example, 75 percent of respondents say defect tracking tools and 69.4 percent say source code control tools are important to DevOps. In contrast, 41.4 percent chose application performance management tools as being important and 44.7 percent chose release management tools, which are more operationally focused. The results are interesting nonetheless.
Looking at the consolidated charts, it seems that DevOps is providing a benefit in terms of easing software upgrades, reducing the number of defects that make it into production and releasing software upgrades faster. Yet when broken down, the results tell a different story. Replay Solutions supplied, at our request, charts that indicate the performance of organizations without any DevOps. Combined, 299 respondents with a dedicated DevOps staff of shared duties reported that less than 10 percent of defects made it into production, compared with 192 respondents that have no DevOps.