Kernel-based monitoring of container activity in a compute environment
An illustrative agent deployed in a compute environment monitored by a data platform is disclosed. The agent may access kernel data generated by a kernel of an operating system used within the compute environment. Based on the kernel data, the agent may detect a launch of a new container entity of a set of container entities deployed to the compute environment and managed by a container runtime executing on the operating system. Based on the detecting of the launch of the new container entity, the agent may then provide, to the data platform, agent data that indicates the launch of the new container entity. Corresponding methods, systems, and products are also disclosed.
1 . A method comprising:
accessing, by an agent deployed in a cloud compute environment monitored by a data platform, kernel data generated by a kernel of an operating system used within the cloud compute environment;
detecting, by the agent and based on the kernel data, a launch of a new container entity of a set of container entities deployed to the cloud compute environment and managed by a container runtime executing on the operating system; and
based on the detecting, providing, by the agent to the data platform, agent data that indicates the launch of the new container entity, wherein:
the kernel data includes a creation callback generated by the kernel of the operating system; and
the detecting of the launch of the new container entity includes:
identifying, within the creation callback, a container entity name, and
determining that the container entity name has not yet been observed in previous callbacks produced by the kernel of the operating system.
2 . The method of claim 1 , wherein the new container entity is implemented as a containerized application or as a pod that includes one or more containerized applications.
3 . The method of claim 1 , further comprising:
collecting, by the agent and subsequent to the detecting of the launch of the new container entity, additional data associated with the new container entity, the additional data collected from the container runtime based on a container entity name indicated by the kernel data; and
as part of the providing of the agent data to the data platform, providing the additional data collected from the container runtime.
4 . The method of claim 1 , further comprising:
detecting, by the agent and based on additional kernel data generated by the kernel subsequent to the detecting of the launch of the new container entity, a launch of a new process on the operating system;
correlating, by the agent and based on a container entity name indicated by the kernel data and a process name indicated by the additional kernel data, the new process to the new container entity; and
as part of the providing of the agent data to the data platform, providing process data indicating that the new process has launched and is correlated with the new container entity.
5 . The method of claim 1 , further comprising:
collecting, by the agent from the container runtime, container data indicative of one or more container entities of the set of container entities that are deployed to the cloud compute environment and managed by the container runtime prior to the launch of the new container entity; and
as part of the providing of the agent data to the data platform, providing the container data collected from the container runtime.
6 . The method of claim 1 , further comprising:
identifying, by the agent, a set of processes executing on the operating system and associated with the set of container entities;
monitoring, by the agent, network activity produced or received by the set of processes that has been identified; and
as part of the providing of the agent data to the data platform, providing process data generated based on the monitoring of the network activity.
7 . The method of claim 1 , wherein the agent data provided to the data platform is configured to be used by the data platform for generation of a data model that facilitates the data platform in assessing a security posture of the cloud compute environment.
8 . The method of claim 7 , wherein:
the data model is a graph comprising a plurality of graph nodes connected by a plurality of edges;
each graph node of the plurality of graph nodes represents a logical entity and each edge of the plurality of edges represents a behavioral relationship between graph nodes connected by the edge; and
the plurality of graph nodes includes a container launch event graph node associated with the launch of the new container entity to allow the assessing of the security posture to account for the launch of the new container entity.
9 . The method of claim 7 , wherein the assessing of the security posture of the cloud compute environment includes identifying, based on the data model, at least one of an anomaly, a security threat, or a compliance issue with respect to the set of container entities deployed to the cloud compute environment.
10 . A computer program product embodied in a non-transitory computer-readable storage medium and comprising computer instructions for performing a process comprising:
accessing kernel data generated by a kernel of an operating system used within a compute environment to which the computer program product is deployed;
detecting, based on the kernel data, a launch of a new container entity of a set of container entities deployed to the compute environment and managed by a container runtime executing on the operating system;
based on the detecting, providing, to a data platform that monitors the compute environment to which the computer program product is deployed, agent data that indicates the launch of the new container entity;
detecting, based on additional kernel data generated by the kernel subsequent to the detecting of the launch of the new container entity, a launch of a new process on the operating system;
correlating, based on a container entity name indicated by the kernel data and a process name indicated by the additional kernel data, the new process to the new container entity; and
as part of the providing of the agent data to the data platform, providing process data indicating that the new process has launched and is correlated with the new container entity.
11 . The computer program product of claim 10 , wherein:
the kernel data includes a creation callback generated by the kernel of the operating system; and
the detecting of the launch of the new container entity includes:
identifying, within the creation callback, a container entity name, and
determining that the container entity name has not yet been observed in previous callbacks produced by the kernel of the operating system.
12 . The computer program product of claim 10 , wherein the process further comprises:
collecting, subsequent to the detecting of the launch of the new container entity, the additional kernel data associated with the new container entity, the additional kernel data collected from the container runtime based on a container entity name indicated by the kernel data; and
as part of the providing of the agent data to the data platform, providing the additional kernel data collected from the container runtime.
13 . The computer program product of claim 10 , wherein the process further comprises:
collecting, from the container runtime, container data indicative of one or more container entities of the set of container entities that are deployed to the compute environment and managed by the container runtime prior to the launch of the new container entity; and
as part of the providing of the agent data to the data platform, providing the container data collected from the container runtime.
14 . The computer program product of claim 10 , wherein the agent data provided to the data platform is configured to be used by the data platform for generation of a data model that facilitates the data platform in assessing a security posture of the compute environment.
15 . The computer program product of claim 14 , wherein:
the data model is a graph comprising a plurality of graph nodes connected by a plurality of edges;
each graph node of the plurality of graph nodes represents a logical entity and each edge of the plurality of edges represents a behavioral relationship between graph nodes connected by the edge; and
the plurality of graph nodes includes a container launch event graph node associated with the launch of the new container entity to allow the assessing of the security posture to account for the launch of the new container entity.
16 . A computing system within a compute environment to which an agent is deployed, the computing system comprising:
a memory storing instructions implementing the agent; and
one or more processors communicatively coupled to the memory and configured to execute the instructions to perform a process comprising:
accessing kernel data generated by a kernel of an operating system used within the compute environment;
detecting, based on the kernel data, a launch of a new container entity of a set of container entities deployed to the compute environment and managed by a container runtime executing on the operating system; and
based on the detecting, providing, to a data platform that monitors the compute environment, agent data that indicates the launch of the new container entity, wherein:
the kernel data includes a silo creation callback generated by the kernel of the operating system; and
the detecting of the launch of the new container entity includes:
identifying, within the silo creation callback, a container entity name, and
determining that the container entity name has not yet been observed in previous callbacks produced by the kernel of the operating system.
17 . The computing system of claim 16 , wherein the process further comprises:
detecting, by the agent and based on additional kernel data generated by the kernel subsequent to the detecting of the launch of the new container entity, a launch of a new process on the operating system;
correlating, by the agent and based on a container entity name indicated by the kernel data and a process name indicated by the additional kernel data, the new process to the new container entity; and
as part of the providing of the agent data to the data platform, providing process data indicating that the new process has launched and is correlated with the new container entity.
18 . The computing system of claim 16 , wherein the agent data provided to the data platform is configured to be used by the data platform for generation of a data model that facilitates the data platform in assessing a security posture of the compute environment.
19 . The computing system of claim 18 , wherein:
the data model is a graph comprising a plurality of graph nodes connected by a plurality of edges;
each graph node of the plurality of graph nodes represents a logical entity and each edge of the plurality of edges represents a behavioral relationship between graph nodes connected by the edge; and
the plurality of graph nodes includes a container launch event graph node associated with the launch of the new container entity to allow the assessing of the security posture to account for the launch of the new container entity.