Productivity improvements in software development and deployment strategies will ultimately have to reckon with the lingering lack of feedback between design time and run time.
Software is still a hand-off affair, with developed applications getting tossed into production with little collaboration between the builders and the operators -- before or after the hand-off.
Things could and should be different. Thanks to search-based technologies and services now entering the market, we may be on the verge of a new productivity boomlet that leverages more needles from more haystacks.
With proper access to information about how code actually behaves in real-world use, developers could better produce reliable applications and infrastructure. Architects and systems operators could better anticipate how to meet the demands and service level agreements (SLAs) of quickly provisioned applications if they had greater visibility into the hit on resources -- and potential disruptions -- from newly minted applications and services. Virtualization will only exacerbate the deployment complexities.
Wouldn't it be beneficial then if the information about what goes on "on the other side" were made available proactively to each side of the equation? Search functions applied directly to both sides of the development and deployment fence would allows those open a bright new window into what remains murky and mysterious from the outside.
Developers with proper access to indexes and meta data could use search to quickly find highly specific information about run time environments and stacks as they write and test their code, and as they seek out the best components, objects and methods suitable for specific runtime scenarios.
Conversely, operators faced with slow-running applications -- or worse -- could search into the source code for clues about root causes of glitches, and much easier and faster identify and remediate the problems. They could also clearly point out the impactful issues back to the development and test teams, to prevent the glitch from recurring in the future.
We're already seeing a great deal of value from operations-side search, and the extension of that value due to platform approaches, APIs and open collaboration. Splunk is providing a path on IT search as a platform. [Disclosure: Splunk is a sponsor of BriefingsDirect podcasts, including this one on Splunk Base.]
Other vendors are emerging to fruitfully employee search to the source code management (SCM) space, such as Krugle. Krugle offers a search benefit for open source assets, as well as enterprise development assets.
Now when we can jibe software characteristics and collaborate across the design time-run time divide based on the type of insight that the likes of Splunk and Krugle provide, well then we'll be in a better software age. It may be sooner than we think.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment