Availability Metrics
Synthetic Transactions are designed to isolate client application or data model/data condition related issues from basic cluster health issues. They are not meant to be impacted by client or data issues unless they get to the point of impacting the general health of the cluster.
- Synthetic Transaction Latency: Write (ms) – Average write latency for 20 writes.
- Synthetic Transaction Latency: Read (ms) – Average read latency 20 reads.
- Synthetic Transaction Latency: Read (ms) – Number of write request errors.
- Synthetic Transaction Error: Read (count) – Number of read request errors.
- Health Check Error: Keeper (count) – Number of ClickHouse Keeper errors.