- Change start_at from 'end' to 'beginning' for complete log history
- Fix transform configuration to match authentication service
- Add k8s_cluster receiver to collect container stdout logs
- Remove problematic json_parser operator
- Ensure consistent log processing across services
- Remove k8scluster receiver (invalid type name)
- Remove unused otlp receiver definitions
- Keep only filelog receiver which is actually used in pipeline
- Fixes CrashLoopBackOff issue in central-storage pod
- Replace Chinese comments with English in opentelemetry.yaml files
- Maintain consistent English-only comments in freeleaps-ops repository
- Keep the same functionality while improving code readability
- Change receivers from [filelog, otlp, k8scluster] to [filelog] only
- Prevent duplicate logs from multiple collection sources
- Keep only filelog receiver to collect logs from application log files
- This eliminates duplicate logs appearing in Grafana for devsvc and central-storage
- Change start_at from beginning to end to avoid historical log duplication
- Add poll_interval: 1s for real-time file monitoring
- Optimize batch processing: send_batch_size: 1, timeout: 1s
- Add PVC template for log persistence to reduce log loss risk
- Update deployment to support persistent volume for logs
- Introduced log ingestion configuration in values files for centralStorage, content, notification, and payment services.
- Updated deployment templates to conditionally include OpenTelemetry annotations and volume mounts based on log ingestion settings.
- Added OpenTelemetry RBAC configurations for service accounts and cluster roles to enable logging.
- Implemented OpenTelemetry collector configuration to process logs and export them to Loki.
- Ensured compatibility with existing Helm chart structure and maintained backward compatibility for services without log ingestion enabled.
Signed-off-by: zhenyus <zhenyus@mathmast.com>
