Reviewed on: 2020-11-16
Archivist collects component lifecycle events, inspect snapshots, and log streams on Fuchsia, making them available over the fuchsia.diagnostics/ArchiveAccessor
protocol.
This project is included in the bringup
build product and most others as a result.
The production Archivist is “mounted” in the component topology in the bootstrap realm.
Tests run by run-test-component
have an Archivist embedded in the test realm when either fuchsia.logger/LogSink
or fuchsia.logger/Log
is requested.
Realms can run their own Archivist by running meta/archivist-for-embedding.cmx
from the archivist
package. This has a number of sharp edges today and the Diagnostics team recommends consulting with us in the process of writing a new integration.
Unit tests are available in the archivist-tests
package.
Integration tests for system logging are available in these packages:
archivist-integration-tests
archivist-integration-tests-v2
test-logs-from-crashes
The entrypoint is located in src/main.rs
, with the rest of the code living in src/*.rs
files. Unit tests are co-located with the code and integration tests are located in the tests/
directory.
Each data type the Archivist supports has a directory:
src/inspect
src/lifecycle
src/logs
Archivist accepts a configuration file to allow product-specific configuration separately from its package. The configuration file must be valid JSON with a single object at the top level. Its path within Archivist's namespace is defined by the --config-path
argument.
See the generated docs for Archivist's serde-based config parser for details.