Scylla Metric Update - Scylla 2.3 to 3.0

Scylla 3.0 Dashboards are available as part of the latest Scylla Monitoring Stack

The following metrics are new in Scylla 3.0

  • scylla_cql_filtered_read_requests
  • scylla_cql_filtered_rows_dropped_total
  • scylla_cql_filtered_rows_matched_total
  • scylla_cql_filtered_rows_read_total
  • scylla_cql_reverse_queries
  • scylla_database_large_partition_exceeding_threshold
  • scylla_database_multishard_query_failed_reader_saves
  • scylla_database_multishard_query_failed_reader_stops
  • scylla_database_multishard_query_unpopped_bytes
  • scylla_database_multishard_query_unpopped_fragments
  • scylla_node_operation_mode
  • scylla_storage_proxy_replica_cross_shard_ops

The following metric was updated from Scylla 2.3 to Scylla 3.0

  • scylla_reactor_cpu_busy_ns changed to scylla_reactor_cpu_busy_ms
  • scylla_reactor_cpu_steal_time_ns changed to scylla_reactor_cpu_steal_time_ms
  • scylla_scheduler_time_spent_on_task_quota_violations_ns changed to scylla_scheduler_time_spent_on_task_quota_violations_ms
  • scylla_transport_unpaged_queries changed to scylla_cql_unpaged_select_queries