Released on 05 Dec 2021
- Alerts can be generated from matching logs. The queries to identify matching logs is pre-defined in a rules file. Episilia will push metrics for all matching logs to Prometheus. Alert rules can be setup in Prometheus to generate alerts into AlertManager of Prometheus and notified via all the supported channels such as PagerDuty, Slack, Email etc.
- Metrics for index warming when a search server is started.
- Search queries now support additional special characters such as double quotes, square brackets etc.
- Optimizer server was running into Out Of Memory situations for large messages. This is now fixed.
Configuration for Alert server and Tail logs. The consumer groups specified here should be different from the primary indexer kafka group.
- kafka.group.logwatcher Kafka consumer group to ingest logs when both tail and alert modes are enabled
- kafka.group.logwatcher.tail Kafka consumer group to ingest logs when only tail mode is enabled
- kafka.group.logwatcher.alert Kafka consumer group to ingest logs when only alert mode is enabled.
- kafka.topic.alert.response.out Kafka topic to publish alert data to Prometheus. For internal use.
- alert.rules.file.url s3 path for alert rules files
- alert.prometheus.gateway Prometheus push-gateway url