Logging is a pivotal observability aspect in LitmusChaos as it allows the user to track the exact system behavior during the scenario of a chaos. The logs can be classified into one of the following:
- Litmus Checker Logs: Logs generated as part of the validation for chaos resources that are required to execute a chaos experiment.
- Experiment Logs: Logs generated as part of the steps performed during the chaos experiment, including pre-chaos check logs, chaos injection logs, chaos probes logs, and post-chaos check logs.
- Non-Chaos Workflow Step Logs: Logs generated as part of the workflow steps that facilitate the execution of the chaos experiment, such as chaos experiment installation step logs, chaos revert step logs, etc.
- User requests the logs for any particular workflow step using the ChaosCenter.
- The request for the logs is received by the Backend Server and is forwarded to the Subscriber.
- The subscriber checks if the workflow step is a Chaos Experiment step or not.
- If the workflow step is a Chaos Experiment step, then the Litmus Checker logs and the Chaos Experiment Logs are fetched from the ChaosEngine CR by the subscriber. Else, the logs of the workflow step pod is fetched from the respective workflow step pod by the subscriber.
- The fetched logs are returned to the Backend Server, which returns them to the ChaosCenter.