Our Expertise

Securing the Serverless

Chaitanya Jawale
Posted by Chaitanya Jawale on March 27, 2020

In one of my earlier posts, serverless isn’t exactly serverless, I wrote about how serverless, an operational model in cloud computing, is helping organizations in outsourcing the routine tasks to remote pro teams. With serverless, one can focus more on the features that purely work on user satisfaction instead of administering a complex infrastructure. You can save time, money, and resources with serverless technology deployed in the right manner. However proper security measures while going serverless is a very important aspect. Now, serverless involves applications hosted by or on third-party cloud providers. These applications and the providers already have security checks and measures in place. But in the current scenario, you have to have additional security in place if you value your applications and customers using them.

Serverless security is a shift in the approach organizations view the security of their applications. Organizations must build security around the functions within the applications that are hosted by third-party cloud developers. The additional layer of security ensures application hardening. Altogether adopting serverless diminishes the security-related concerns that are related to managing your servers and resources as most of the responsibility is handled by your cloud provider. Still, there are certain challenges you might face while adopting serverless on the security front. As the structure of serverless application changes, challenges keep on arising. Here are some of the highlighting ones:

  • Visibility: As the amount of information and number of resources increases with serverless the ability to comprehend the data such as everyday event logs, gathering intelligence for the right observability becomes difficult. Besides, serverless apps utilize various services from different cloud providers from several versions and regions. To know your attack points and possible risks, you should be aware of your complete serverless infrastructure. As the app grows it can become very challenging to build as well as maintain this security-focused structure.
  • Protocols, vectors, and attack points multiply to every function: These are the most potent points of attack. A distinctive approach is required for Google Functions, Azure Functions, and AWS Lambda security.
  • Additional Resources: With additional resources, you get more permissions to manage. This leads to creating challenges when it comes to determining permissions for all the interactions.

Serverless provides a prodigious opportunity in speeding up application development by reduced operating costs and expanded agility. And at the same time, it lessens the security concerns around revolving servers and upraises the ones that are related to applications. The areas like operating system, web server, software elements, and programs get shifted to a cloud provider. As a result, the focus should be shifted on moving data, permissions, and directing weak application dependencies. These vulnerable situations can be addressed through the best security practices. Practices like not trusting any input, failing securely, and following the principle of least privilege tops the list. To make your future decisions secure, here are some of the best practices that you can follow:

  • Patch your function dependencies - Application dependency is normally widespread and the most vulnerable part and serverless, because of the third-party dependencies, make it more challenging. As an application owner, you are responsible for patching the libraries during a vulnerability.
  • Follow the least privilege principle - Deploy your functions and keep the permissions set to a minimum. This will reduce the attack surface.
  • Preserve and isolate function perimeters - Apply perimeter security at the function level itself especially when multiple functions are deployed to create a whole aggregative workflow. This will help to keep other functions unaffected even if one function becomes vulnerable.
  • Clean up the event input to avoid any kind of injection - When harmful data manages to bypass a data sanitization and reaches the event payload, it might lead to susceptible injection attacks. This data should be handled properly to make sure the processing functions are safe from attacks.
  • Use API gateways for a security buffer - You can expose functions through API gateways which is considered a good practice. It works as reverse proxies and is good at providing a separation layer between users and functions.
  • Follow secure coding conventions - Activities like sessions for code review to identify application code security issues early in the software development process is a must. To avoid vulnerable application code, your developers should follow the secure coding conventions rigidly.
  • Secure and check the data in movement - Data exchange between all the functions and interdependent services should be closely monitored. Security extension protocols and certification verification processes should be implemented on priority.

Monitoring and logging are the pillars of any system security and serverless is no exception to this. You can easily avoid breaches as well as look into the intruder’s actions with detailed logging. Practices like input validation, and regularly checking the dependencies of external libraries and web resources to keep a track of vulnerabilities are highly advisable. Combine efficient monitoring with measures such as proper API gateway, restricted function permissions, encryption, application firewall, behavioral analysis tools & IoT botnet solutions will bolster your serverless security.

Topics: Cloud, DevOps, & Containers, Security, Serverless, SecOps, Serverless Security

Leave Comment

Subscribe Email

    Post By Topic

    See all