Crowdstrike Falcon Disruption: Why SaaS Security Vendors Need to Focus on Designing for Failure

Executive Summary :

• Systems should be designed for failure – this is an Indusface core tenet 
• Business continuity is a shared responsibility between vendors and customers 
• Indusface not affected by this outage 

By now, everyone is aware of the CrowdStrike Falcon update that caused major disruptions to key services such as hospitals, flights, news channels, and millions of end-user and cloud-based Windows machines worldwide. Details, including recovery options, can be found in the CrowdStrike advisory. 

At Indusface, designing for failure is a core tenet that we use while building all our SaaS products. After all, despite the best processes, checks, and balances, any system can fail. No system or process including ours is foolproof.

Given that Indusface AppTrana Web Application & API Protection (WAAP) platform is an edge security solution, customer’s assets are only accessible through our systems. In the event of catastrophic unavailability of our core systems, we can rapidly and automatically transition them to an independent system. This switchover is granular, down to a single asset, customer region, or entire system, ensuring the availability of the site even in the event of a failure.

More details on this can be found  here. We ensure that we take into consideration the availability needs vs. security needs of the customer and provide them with the option to choose how they react if there is a need, i.e., should we fail open vs. fail close. This is one example of how we design our systems to reduce impact during eventualities.  

We follow the same principle in every design decision. We build our systems in a way that we react to failures in the least disruptive way possible. At various levels in our architecture, we have fail-safes and mechanisms in place to ensure gradual, localized failure so that impact is minimized when inevitable issues happen. It is because of this that we can provide a 100% availability guarantee to our customers. 

To talk about lessons from this outage, this could have been prevented on two counts, 1, had the software been designed in such a way that there was a plan B in case of a failure on the update and 2, the businesses in question designed processes in such a way that they had backups systems to rely on in case of outages on the primary systems.  

The responsibility of business continuity therefore lies as much on software vendors as on businesses. 

Indusface systems were not affected during the recent CrowdStrike outage, and all our services are up and providing optimal protection to our customers’ websites. We stand in solidarity with our tech community and are happy to help our customers and the community at large in any way we can. As the world recovers from the sudden shock, we urge organizations and vendors to revisit their architecture and redesign their systems and processes assuming things that can fail will eventually fail at some point. 

Stay tuned for more relevant and interesting security articles. Follow Indusface on Facebook, Twitter, and LinkedIn.