In their current state, logs only meet the basic requirement for debugging purposes. Logs mainly allow basic monitoring of communication messages between components and are split in four categories:
- middleware - broker: encompases messages sent or received between a middleware and the broker, for example:
A Middleware logs it's view of a brokerage request to the Broker:
middleware-client:10000 - Requesting brokerage of contract
- service/client - middleware: encompases messages exchanged between agents participating in an execution and their corresponding middlewares. Examples of these events are:
A Middleware logs it's view of a message sent by a process running behind it:
Received CardDetailsWithTotalAmount: {232 22 22 110}
- middleware - middleware: encompases the logging events associated to messages exchanged between middlewares, for example:
A Middleware logs when it dispatches a message to a another middleware hosting a remote participant:
Sent message to remote Service Provider for channel cf4ea1f5-44d3-4b27-b489-64c424911611, participant PPS
- critical internal actions of the infrastructure's component: these messages are used for monitoring the behaviour of the components by following their execution by tracking chosen internal actions, for example:
A Middleware logs when it receives a first outbound message from a participant:
Received first outbound message to send on channel cf4ea1f5-44d3-4b27-b489-64c424911611 for participant Srv. Opening connection to remote Service Provider.
A Middleware starts the routine that periodically sends outgoing messages on an outbox buffer:
Started sender routine for channel cf4ea1f5-44d3-4b27-b489-64c424911611, participant Srv
Several debugging entries of logs can be found in the code. These entries are used to get a close look at critical parts of the code implementing the different components, both of the infrastructure, and the example provided with the implementation.
As we mentioned before, SEArch has a basic logging infrastructure based on using the standard output stream of the processes.
When running within a Docker container it is possible to obtain and visualize the logs by using built in capability of the docker compose pluggin.
For a basic usage of this infrastructure a user can try the following commands:
- Visualizing the a time based composition of all the containers:
docker compose logs -f- Visualizing the log of a specific container:
docker compose logs [container_identifier]