In light of the recent Equifax breach, Gene Kim and speakers from the upcoming DevOps Enterprise Summit San Francisco (DOES17) dissected the situation and discussed the technical leadership lessons learned while offering their own expert advice for handling crisis situations. The following are more highlights from the discussion ...
In DEVOPSdigest's first annual list of DevOps Predictions, experts — analysts and consultants, and the top vendors — offer thoughtful, insightful, often controversial and sometimes contradictory predictions on how DevOps and related technologies will evolve and impact business in 2016. Part 4 covers analytics, SaaS, testing and more.
AUTOMATING DEVOPS WITH ANALYTICS
Over the past five years, DevOps teams have focused on automating the delivery of applications into production using provisioning/release/build/config automation tools. It's now time for DevOps teams to automate the way they support applications in production by leveraging software analytics and machine learning to automate the manual tasks of incident detection, troubleshooting and root cause analysis. Relying on humans to manually analyze or search for answers makes no sense in a culture of agility and automation. DevOps teams need to start leveraging machine learning technologies so they can automate much of the tedious tasks that still exist for supporting applications.
Chairman, VP of Product Marketing, Moogsoft
The increasing complexity of enterprise applications has resulted in too many moving parts that impact the quality of service to end users. While monitoring tools successfully exposes more performance metrics than ever, the growing number of charts that visualize symptoms does not make manual root cause analysis via correlation easier any more. DevOps teams will start looking for tools that automatically interpret and analyze performance data, so that they could ensure high SLA while spending most of their time on developing new functionality.
Co-founder and CEO, Plumbr
Next year, we will see DevOps teams augment traditional APM tools and rule-based monitoring with behavioral analytics. By complementing the output of already-valuable APM tools with a behavioral analytics solution, previously unknown anomalous behavior can be identified in near real-time, enabling fast moving DevOps teams to remain agile by focusing more on deploying and testing new software and less on sorting through a sea of alerts. With more data being created than ever before, behavioral analytics powered by automated unsupervised machine learning will enable DevOps teams to transform gained insights into a competitive advantage.
VP of Products, Prelert
NEW ANALYTICS TOOLS FOR TRACKING CHANGES
While DevOps and other agile methodologies already allow IT to deliver changes into production at an overwhelming pace, we see how this is also introducing real challenges from a stability perspective, leaving IT ops vulnerable. Although most performance incidents ultimately result from changes (about 85% of incidents can be traced to changes, according to Gartner), monitoring and analyzing changes is still overlooked by DevOps and Ops teams. We expect to see a new generation of analytics-powered DevOps and ITOA tools that will help deal with the challenges introduced by DevOps by collecting and analyzing all actual changes across entire environments, assessing their impact and thus allowing IT teams to plan accordingly. Monitoring all actual changes and analyzing them together with release and deployment data, will allow for smooth, error free releases. Specifically, these analytics tools will be able to identify key gaps between pre-production, production and DR; detect inconsistency and missing pre-requisites that impede deployments; validate deployments by accuracy and consistency; identify root-causes of stability issues due to deployments. IT won't have to sacrifice stability for agility anymore.
BUSINESS INTELLIGENCE VITAL TO DEVOPS SUCCESS
Analytics will focus on both IT and Business analytics. The first major implications will be a decline in monitoring tools and more correlation between existing tools. Major companies will reliaze that they have all the tools they need but they don't connect them correctly. Business analytics is a large market but very immature. The rise of DevOps will help business analytics focused companies grow.
Online Performance Consultant and Founder of Blue Factory Internet
DEVOPS LEVERAGES SAAS
Over the past few years DevOps has become an accepted approach and philosophy for building and managing today's systems. My prediction is that 2016 will see a lot of mainstream adoption of DevOps, in particular in the large enterprise. This means more adoption of technologies that make up the modern DevOps toolkit, primarily SAAS services. Over the past few years, people have really started to move away from the on-prem, "roll your own" solutions, and some of the old dinosaurs that tried to provide everything in one box. Now, DevOps professionals are more often taking advantage of specific, cloud-based services that are more flexible, require less investment up front, and practically zero management.
Senior Director, Log Management & Search, Rapid7
In 2016 the adoption of cloud delivered SaaS offerings will begin to take hold in areas that, especially for the larger more established enterprises, have so far been largely off limits. Long held assumptions concerning the necessity to host applications on-premises such as those responsible for infrastructure management will be challenged as the demonstrable security and reliability of cloud hosting services becomes broadly accepted. The industry sectors that are assumed to be unable to take advantage of these software product delivery options will start to experiment and rethink their assumptions leading to a previously unthinkable revolution in how they take advantage of third party services. With SaaS comes the opportunity to benefit from the continuous development and deployment principles of DevOps and those vendors already practicing this methodology internally will be best placed to exploit this exciting new frontier.
Principal Solutions Architect, Entuity
CONSOLIDATION OF AUTOMATION TOOLS
Automation is proven to enable DevOps speed and efficiency. In 2016 we'll start to see IT organizations simplify automation with tool chain consolidation. Infrastructure automation and application stack automation tools will start to merge, and IT organizations will look to use one tool that works across data center, private cloud and public cloud environments.
VP of Product Marketing, CliQr
PERFORMANCE TESTING BECOMES CRITICAL COMPONENT OF DEVOPS
In 2016, performance testing will come to be viewed as a first class component of a continuous delivery and DevOps pipeline, and as being essential to overall software quality practices. This will happen through the adoption of emerging open sources tools that provide the right capabilities for modern performance testing automation. Developers and DevOps teams will be able to implement performance testing patterns and practice more easily, and better drive overall quality initiatives in their organizations.
Chief Evangelist, BlazeMeter
2016 will be the year where we start seeing an increased focus on app testing as a critical part of the application development lifecycle. Access to cloud-based, scalable and turnkey testing tools that integrate well with continuous delivery flows will be paramount. The DevOps tools conversation will expand from continuous integration and infrastructure configuration solutions to a broader set that includes scalability testing and performance testing, as well as unit testing. Seamless testing is critical for any DevOps flow to succeed, for the DevOps promise to be delivered and for businesses to achieve the right speed of execution.
Founder and CEO, Nouvola
LOAD TESTING MOVES TO THE CLOUD
In 2016, we'll see more enterprises moving load testing to the cloud. Companies are under constant pressure to deliver products and services to the market faster than ever before. Meanwhile, business owners are exhausted by in-house IT departments with slower than needed delivery cycles. In this dichotomy, performance must not be sacrificed. By using load testing from the cloud capabilities, the business and technology teams both get what they need: Technology gets an affordable, secure way to test system and application performance before it reaches production, and the business doesn't have to wait so long that it loses competitive advantage.
Chief Technology Evangelist, Hewlett Packard Enterprise
DEVOPS APPLIED TO INFRASTRUCTURE DELIVERY
DevOps moves out of delivering skunkworks and front office applications toward the back office, taking over infrastructure delivery for most customers. Gaining momentum inside more traditional enterprises, DevOps has been proving its worth as a way to replace slow and inefficient service delivery. The time is right for agile processes to completely take over other areas, and make IT infrastructure delivery faster, more reliable and more accountable.
Director of Product Marketing, SevOne