What we achieved

  • Zero Configuration deployment process

  • Leveraged existing processes in the SDLC

  • Detailed security vulnerability records

To who

  • Industry: Information Technology

  • Products: HCL AppScan

  • Region: North America/US

Overview

  • Part 1

    Challenge

    Our customer was faced with the following business challenges:

    Improving the security protection of their products without disrupting the current SDLC process.

    Reducing the probability of a security issue that could delay shipping of new versions.

  • Part 2

    Solution

    Integrate IAST into the customer’s existing QA process and leverage automatic, manual and sanity tests to extend Application Security Testing (AST) coverage and transform DevOps to DevSecOps.

  • Part 3

    Results

    Improved AST coverage and remediation processes, due to informative records of security issues such as full call stacks and exploit examples that are reported by the IAST agent.

The Challenge

Business Case for IAST

The company was already utilizing DAST as part of their SDLC, mostly in the late stages. This common practice provided good results, but had several downsides to it:

  • When a significant security vulnerability was discovered, it caused a delay in the release, since DAST was introduced as one of the last steps before a new version was shipped. Remediation efforts for security vulnerabilities were high due to the DAST scanner's less detailed information.
  • There was a significant time gap between writing the code and discovering vulnerabilities.

We were surprised by the deployment process. We were expecting something more complicated than deploying a WAR file to our Tomcat!

Technical Manager DevOps team

The Solution

Integrating IAST

The company has an extensive Quality Assurance (QA) process due to its codebase's size and complexity. The QA process includes automated and manual testing that ranged from simple sanity scenarios to complicated edge cases. Every new version also added more functionality, so further tests was introduced into the QA process.

The QA infrastructure is Docker-based and orchestrated using Jenkins. Since the team didn't want to change their existing containers, they decided to integrate IAST by using a simple script that utilizes AppScan's APIs to download and deploy the agent to the web server, after applications are successfully built and published.

The amount of information I receive per issue is beneficial for the prioritization and remediation process.

System Architect

The Results

Effects

A significant benefit that developers instantly reported was the amount of information the security vulnerabilities contained. Having the line of code that originated the issue, along with an example of an exploit that triggered it, reduced remediation efforts significantly. Since the QA process is adjacent to the development process, the code changes that resulted in new security vulnerabilities are fresh in developers minds when approaching to resolve security issues.

Another benefit that the security team reported was reducing issues detected in DAST scanning, since the QA process now helped to resolve issues earlier in the SDLC.

From a maintenance perspective, the Security and DevOps teams were impressed since integrating the IAST agent only requires a single straightforward script, and the agent itself is evergreen (meaning that it updates automatically). Another great thing is that the QA team can keep adding new tests for every new functionality it develops, keeping AST coverage up to date with every new version. The process keeps improving as a byproduct of the SDLC itself.

About the company

Due to the cybersecurity domain's sensitive nature, the company requested to stay anonymous in this particular case study. The company is a software company in the IT e market that provides services to SMBs and large enterprises.

The technology stack used in this case study is:

  • Java
  • Tomcat
  • Docker
  • Jenkins

Related Capabilities