You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: CHANGELOG-6.md
+38Lines changed: 38 additions & 0 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -5,6 +5,44 @@ All notable changes to this project will be documented in this file.
5
5
The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
6
6
and this project adheres to [Semantic
7
7
Versioning](http://semver.org/spec/v2.0.0.html).
8
+
## [Unreleased]
9
+
10
+
### Added
11
+
- Added backend-side validation for `runtime_assets` during CheckConfig creation and update.
12
+
- If any referenced asset does not exist in the namespace, the request will be rejected with a validation error.
13
+
- This change improves usability by catching missing or invalid assets early, preventing runtime execution failures and reducing debugging overhead.
14
+
- Added `ttl_status` field to `Check` and `CheckConfig` to configure the status (warning or critical) for TTL failure events. This field accepts values 1 (warning) or 2 (critical), with critical (2) as the default to maintain backward compatibility.
15
+
- Added asset lifecycle management with `last-accessed` timestamp tracking to enable intelligent cleanup of unused assets based on usage patterns.
16
+
- Added plugin-based hook system for agent extensions with `BeforeAgentRun` and `AfterAgentRun` lifecycle hooks, supporting priority-based execution ordering and configurable error handling.
17
+
- Added `assets-cleanup-interval` and `assets-cleanup-max-age` configuration options for controlling periodic asset cleanup behavior.
18
+
- Added `FindUnusedAssets()` and `DeleteAsset()` functions for programmatic asset cleanup operations.
19
+
- Added `GetAPIRouter()` and `GetDB()` methods to agent for plugin access to HTTP router and database connection.
20
+
- Added `continue_on_error` field to `core/v2.Pipeline`.
21
+
This flag controls whether pipeline execution should continue when an error occurs in a handler, filter, mutator, or during asset resolution.
22
+
Defaults to `false`, preserving existing behavior.
23
+
- Added `ttl_status` field to `Check` and `CheckConfig` to configure the status (warning or critical) for TTL failure events. This field accepts values 1 (warning) or 2 (critical), with warning (1) as the default to maintain backward compatibility.
24
+
- Added Prometheus metrics collection for HTTP API requests, including request count (`sensu_go_http_requests_total`), request duration (`sensu_go_http_request_duration_seconds`), and client error count (`sensu_go_http_client_errors_total`) for better observability of backend API performance.
25
+
26
+
### Changed
27
+
- Improved asset error handling with new `AgentAssetError` interface providing structured error reporting with exit codes and detailed messages.
28
+
- Enhanced asset manager architecture by restructuring initialization to support pre-initialized database connections and better enterprise integration.
29
+
- Improved command execution timeout error handling for more reliable test execution.
30
+
31
+
### Fixed
32
+
- Fixed asset retrieval error aggregation to continue processing remaining assets when individual assets fail, using `multierr` for comprehensive error reporting.
33
+
- Upgraded Go version to 1.24.3 across all CI/CD pipelines
34
+
- Updated CircleCI, AppVeyor, and GitHub Actions to use Go 1.24.x
35
+
- Updated minimum Go requirement in README from 1.16 to 1.24
36
+
37
+
### Fixed
38
+
- Resolved Go 1.24 compatibility issues in agent version and event tests
39
+
- Fixed non-constant format string error in pipeline filter legacy code
40
+
- Updated `GetAssets` method to return an error when one or more required assets are missing.
41
+
When `continue_on_error` is `false`, missing assets or other errors will stop the pipeline execution (existing behavior).
42
+
When set to `true`, the pipeline will continue executing subsequent workflows until one succeeds.
43
+
- Modified handler execution behavior to return a non-zero status code on failure.
44
+
Previously, handler errors were only logged without affecting pipeline.
0 commit comments