Today’s application-driven economy has fuelled an uptake of DevOps practices; today, more than a quarter of average IT departments will be made up of members of a DevOps team. Tasked with continuously delivering ever higher quality applications at faster speeds, these teams offer many advantages for business growth. The global DevSecOps market is predicted to reach $6.1 billion by 2023, growing at a CAGR of 33 percent.

The increasing importance of ensuring application security has seen the function evolve. Up to 78 percent of organisations now claim to have adopted a DevSecOps function to some degree, with developers, IT operations, and security specialists working closely together to continuously develop and deliver secure applications and services quicker than ever before.

Michael Segal, Area VP of Strategy, NETSCOUT

Facilitating a continuous delivery pipeline can often require DevSecOps teams to build, test, integrate and deploy several new releases a day. But, as this deployment pipeline gathers speed, there is an increased risk of potentially harmful flaws and vulnerabilities going unnoticed until it’s too late to avoid customer impact.

To better protect against potential security risks, and avoid customer impact due to delays arising from taking remedial action, DevSecOps require complete end-to-end visibility of the entire service delivery infrastructure, applications it supports, and their respective interdependencies from the moment development begins.

DevSecOpsSecurity from the Start

Analysing data after a breach occurred will certainly help with forensic investigations, but detecting flaws and vulnerabilities in an application in real time during the continuous development and delivery process allows developers, operations, and security teams to deal with any issues more proactively and with considerably more agility.

It’s therefore important that application security requirements are considered as development begins, to identify as many issues as possible early on rather than at the end of the process. If an organisation is using a waterfall development methodology, in which software is deployed on an infrequent basis, it’s possible to manage the impact that any application security vulnerabilities identified during the verification and validation phase may have on the overall process. When DevSecOps teams are deploying new releases on a daily basis, any holdups are unacceptable.

Software development life cycle (SDLC) security requirements, such as the elimination of trap doors, back doors, and covert channels during the initiation and development phases, should be key considerations. Static, dynamic, fuzz and interface testing for the relevant use case should be carried out during the assessment phase; vulnerability assessment and penetration testing combined with Continuous Monitoring, to identify threats and vulnerabilities, should be conducted during the maintenance phase.

With end-to-end visibility of the entire service delivery infrastructure, applications and respective interdependencies throughout the SDLC, DevSecOps can enjoy a common situational awareness that will allow developers to view any given situation, collaborate and take the necessary action as soon as an issue is identified, rather than having to rely on the operations or security team to bring it to their attention after the problem is discovered in production environment.

Smart Data

In today’s connected world, there is an expectation of digital experiences to work seamlessly and securely.  However, the accelerating speed of continuous development, the complexity of hybrid cloud infrastructure and migration to microservices based software architectures makes it very challenging to continuously monitor services by utilizing multitude of domain specific management tools. While each management system provides fine grained views of the individual domain, such as network, application, server or database, it omits the deep insight into the interdependencies across domains which is critical to achieve a common situational awareness for the DevSecOps teams.

Achieving the necessary level of visibility into the interdependencies at the data, network, transport, session, and application layers, requires smart data – metadata based on the processing and organisation of wire data, or IP traffic flows, at its point of collection and optimised for analytics at the highest possible speed and quality. Unlike log data, which needs to be collated and analysed before it can be acted upon, smart data analyses every IP packet that traverses the network during a development cycle and beyond, in real time. The information is used to deliver meaningful and actionable insights. In turn, this creates a common situational awareness for all parties involved in the SDLC.

Smart data also provides actionable intelligence on events as they happen, thereby enabling every team of the DevSecOps function, from developers to operations, and quality assurance to security, to collaborate effectively, each fully mindful of how telemetry, dependencies, and automated feedback are evolving during the development process.

Maintaining Speed and Agility

The innovation and ongoing of success of businesses is dependent on an effective digital transformation. With DevSecOps teams required to produce quality and secure code at speed in order to fully unlock the advantages of such a transformation, there is an element of risk.

To mitigate this risk and avoid potential harm and costly delays in deployment time, security must now be built in from the start of the SDLC process and monitored throughout.  The complete visibility across the service delivery infrastructure, applications and their respective dependencies and the Continuous Monitoring afforded by smart data allows DevSecOps to maintain the speed and agility they need to fulfil their function, and enjoy the common situational awareness they need to assure that potential flaws in performance and security are resolved before they impact end-users.

See also: Where Do You Want to be on the DevOps Maturity Scale?