Add transforms to change data before replay
The Speedscale analyzer examines a set of traffic and outputs the artifacts needed for replay:
- action file - a set of RRPair data prepared for the generator
- reaction file - a set of signatures (request patterns) and responses for the responder
- raw file - a set of traffic matching the filter criteria
In many cases, the process will work out of the box using the
standard tokenizer configuration.
This process can become complex so make sure to join the Speedscale Slack community and we will be happy to walk through your specific use case.
As a basic starting point, the snapshot transform configuration has the following sections:
- generator - applied to all RRPairs in the action file
- generatorVariables - used to store values used by the generator repeatedly - not used by the analyzer
- a Kubernetes secret used to sign JWT tokens
- responder - applied to all RRPairs in the reaction file for the purposes of signature matching
- update the request header to match a variety of values instead of just one
- responderVariables - used to store values used by the responder repeatedly - not used by the analyzer
- an environment variable used by the service under test
- responderResp - applied to responder responses before being sent to the service under test
- a date may need to be changed to the current date instead of when the snapshot was taken
Each transform section is a complete transform configuration as detailed in the Transform Traffic section.