Identifying the root causes of performance issues in Kubernetes (K8s), can often feel like chasing shadows. Real-time traffic visibility provides valuable insights, but it doesn’t always capture the elusive culprits.
By rerecording traffic and making it available for offline investigation, DevOps, SREs, Platform Engineers, and Developers gain a significant advantage in their quest for comprehensive troubleshooting and debugging.
Pattern-based Traffic Recording
Kubeshark Filtering Language (KFL) acts as a filter for L4 streams, giving you control over what traffic gets recorded. You can precisely select the traffic patterns that matter, ensuring that noise and irrelevant data are filtered out effectively.
You can start recording traffic, and store in AWS S3, by providing the following properties in Kubeshark’s config file:
license: FT7YKAYBAEDUY2LD.. your license here .. 65JQRQMNSYWAA= scripting: env: AWS_REGION: us-east-2-this-is-an-example S3_BUCKET: give-it-a-name RECORDING_KFL: "http or dns" # To deactivated remove this field. source: "/path/to/a/local/scripts/folder" watchScripts: true
Get the script and more detailed instructions from here.
Long Term Retention
The recorded traffic is securely uploaded to an AWS S3 bucket dedicated to long-term retention. This ensures that the recorded data remains accessible and available for thorough analysis even after significant time has passed.
The recorded traffic holds valuable insights that can be analyzed over time to uncover hidden patterns and recurring issues.
On-demand Offline Investigation
The true power of recording K8s traffic lies in the ability to conduct offline investigations on-demand.
Use the following command, to investigate the recorded traffic that is stored and retained in the AWS S3 bucket:
kubeshark tap --pcap s3://my-bucket/
The above command initiates Kubeshark’s offline mode, enabling you to explore the contents of the S3 bucket without the need for direct access to your cluster.
The convenience of offline investigation empowers professionals to dig deeper into the recorded traffic, perform comprehensive analysis, and unveil valuable insights for resolving complex issues.
Kubeshark’s dashboard allows you to visualize and explore the recorded traffic using powerful filtering, searching, and analytical capabilities. With this user-friendly interface, you can navigate through the recorded data more efficiently, saving precious time and effort.
RECORDING_KFL property from Kubeshark’s config file to deactivate the recording.
DevOps, SREs, Platform Engineers, and Developers can leverage the ability to record K8s traffic and perform offline investigations to hunt down performance and security culprits with ease.
Traffic recording and offline investigation can lead to faster issue resolution, improved performance, and enhanced security, unraveling the intricate web of interactions within K8s.