fix(monitor): fix Redis Sentinel helpLink, sentinel metric types, add sentinel_timedout_scripts field#4122
Conversation
… timedout_scripts field Three bugs in app-redis_sentinel.yml: 1. helpLink pointed to influxdb_promql docs instead of redis_sentinel docs (copy-paste error). 2. All five sentinel section fields (sentinel_masters, sentinel_tilt, sentinel_running_scripts, sentinel_scripts_queue_length, sentinel_simulate_failure_flags) were declared as type: 1 (string) but Redis INFO sentinel returns integers for all of them. Corrected to type: 0 (number) so dashboards and alerting thresholds work correctly. 3. Added missing sentinel_timedout_scripts field (present in Redis INFO sentinel output, tracks scripts that exceeded the timeout deadline — useful for detecting runaway Lua scripts in HA pipelines). Signed-off-by: abhyudayareddy <54602866+abhyudayareddy@users.noreply.github.com>
Added documentation for monitoring Redis Sentinel instances, including pre-monitoring requirements, configuration parameters, and collection metrics. Signed-off-by: abhyudayareddy <54602866+abhyudayareddy@users.noreply.github.com>
Added detailed monitoring guide for Redis Sentinel, including prerequisites, configuration parameters, and metrics collection. Signed-off-by: abhyudayareddy <54602866+abhyudayareddy@users.noreply.github.com>
|
@Duansg — both documentation files have now been committed to the branch:
|
|
Pls check ci, for related questions, please refer to: code_style_and_quality_guide |
|
@Duansg — I've investigated the two failing CI checks and wanted to share the findings: DOC CI / docs-build (Dead Link Check): The failure is caused by Backend CI / backend-maven-e2e: E2E test failure unrelated to our YAML/docs changes. Required check: The docs requested in your earlier review are now committed:
|
… timedout_scripts field
Three bugs in app-redis_sentinel.yml:
helpLink pointed to influxdb_promql docs instead of redis_sentinel docs (copy-paste error).
All five sentinel section fields (sentinel_masters, sentinel_tilt, sentinel_running_scripts, sentinel_scripts_queue_length, sentinel_simulate_failure_flags) were declared as type: 1 (string) but Redis INFO sentinel returns integers for all of them. Corrected to type: 0 (number) so dashboards and alerting thresholds work correctly.
Added missing sentinel_timedout_scripts field (present in Redis INFO sentinel output, tracks scripts that exceeded the timeout deadline — useful for detecting runaway Lua scripts in HA pipelines).
What's changed?
Checklist
Add or update API