Using Falco to Catch Risky Runtime Behavior Before Merging a Change

Some risks, such as a new dependency opening an outbound connection or a handler shelling out, only appear while code runs and can slip past diff reviews and unit tests. This tutorial uses Falco, the open source Kubernetes runtime security engine and CNCF project, to monitor system calls for a single change before it merges. The changed HotROD route service runs in a Signadot sandbox, and Falco rules scoped to sandbox pods flag unexpected behavior without alerting on normal traffic. In the example, a helper package runs a shell command that reaches an external host on first use, and Falco catches it, causing a Signadot Job to fail until the change is fixed. The tutorial takes about 20 minutes and requires a cluster with the Signadot Operator v1.4.0 or later, the Signadot CLI, and a node kernel that exposes BTF.
This is an AI-generated summary. ShortSingh links to the original source for the complete article.


Discussion (0)
Log in to join the discussion and vote.
Log in