Snapshot Lifecycle
A snapshot is a collection of request and response pairs stored in a file. Snapshots are created by recording a live environment or by importing data from another source like a Postman collection or a Wiremock configuration. On your local machine, you can find your snapshots in the ~/.speedscale/data/snapshots directory. The recorded rrpairs are stored in the raw.jsonl file within the unique snapshot directory. Each line in the file is a flattened JSON object representing a single request and response pair. By editing the raw.jsonl, you are editing the data that is used to generate the mocks.
Your currently active snapshot is shown in the PROXYMOCK pane. Each rrpair is shows as an entry in the tree underneath the hostname or service it belongs to.
If you understand how the git tool works by pushing and pulling files locally to/from the cloud then this system will be familiar.
Snapshots are stored in the ~/.speedscale/data/snapshots directory under a sub folder named after the snapshot's unique identifier. The raw.jsonl file is the main file that contains the request and response pairs. The reaction.jsonl file is the file that contains the configuration for the mock server. The action.jsonl file is the file that contains the configuration for the tests that can be run against your app. There is a metadata file that contains the snapshot's name, description, and other metadata and it is stored in the root ~/.speedscale/data/snapshots directory.
<snapshot UID>.json
<snapshot UID>
├── raw.jsonl
├── action.jsonl
└── reaction.jsonl
Mock Server Lifecycle
Speedscale's mock server uses the outbound (egress) traffic of your application to create a mock server. This process is called "re-analysis" and it only needs to take place when you're ready to use the mocks. The analysis process takes the rrpairs in your raw.jsonl file and creates a file titled reaction.jsonl that configures the mock server with known responses. If you modify the reaction file manually it will work until re-analysis and then be overwritten. Only the raw.jsonl and metadata files are meant to be edited by humans.
Each request received by the mock server is translated into a signature to make it quickly identifiable. If you inspect the reaction.jsonl file you will see that each request is translated into a simplified signature (like a hash map of identifiers) and the response is stored in the http.res JSON object. Modifying the reaction.json file directly is not a good idea because it will be overwritten by the re-analysis process. Instead, find your rrpair in the raw.jsonl file and modify the request or response there.
Updating the Mock Server
When you edit the rrpairs in the tree pane and save a particular rrpair, the raw.jsonl is automatically updated. You must now trigger the re-analysis process to update the reaction.jsonl file. This is done by clicking the Learn button in the PROXYMOCK pane. The mock server will automatically restart.
The rrpair format is not explicitly documented (yet) but it's fairly easy to pick our fields you might want to change. Just make sure to change each occurrence of the field if it appears more than once in the data.
For example, if you want to change the URL of a single rrpair, you can simply edit the location, http.req.url and http.req.uri fields on a single line in the raw.jsonl file. Future versions of proxymock will( likely have a UI for editing the rrpair. Speedscale Enterprise provides a data modification engine called transforms that can be used to modify the rrpair data using automation. Nothing stops you from manually modifying the raw.jsonl in proxymock.
The Learn button is only available when you have an active snapshot.
Changing the Active Snapshot
You can change the active snapshot by clicking the Import button in the PROXYMOCK Editor Command pane. This will open a dialog where you can select a different snapshot. The mock server will automatically restart. Enterprise users can also use proxymock pull to pull a snapshot from another environment.
Merging Snapshots
Combining snapshots is useful if you want to create a mock server that combines data from multiple environments or recordings. On your local desktop you can simply copy/paste two raw.jsonl files together using the text editor. The mock server will automatically re-order the rrpairs based on time stamp when you click the Learn button.
Speedscale Enterprise allows you to merge snapshots from different environments using automated commands and is able to re-order intelligently.
Sharing Snapshots
Snapshots are just files and if you copy/paste the entire directory and metadata to another machine it will work. proxymock does not support being embedded into a CI/CD pipeline but Speedscale Enterprise does. You can learn more about proxymock push and proxymock pull in the Speedscale Enterprise documentation.