29.12.11.5 The replication_applier_metrics Table

This table provides replication applier statistics such as wait times, wait counts, transaction counts, event counts, and byte counts for the applier coordinator.

Note

This table requires installation of the replication_applier_metrics component, which is available with MySQL Enterprise Edition only. See Section 7.5.6, “Replication Applier Metrics Component”, for more information.

The replication_applier_metrics table has these columns:

• CHANNEL_NAME

  The replication channel. See Section 19.2.2, “Replication Channels”.

• TOTAL_ACTIVE_TIME_DURATION

  The coordinator total active time since measurements were restarted. If the applier is stopped and restarted, then all active periods are summed together.

  This information is reset when the replica server is retsarted. It is not reset when the applier stops.

• LAST_APPLIER_START

  The last time (since server start) that the channel's applier was started. This value is reset each time that the applier for the channel named in the CHANNEL_NAME column is started (or restarted).

• TRANSACTIONS_COMMITTED_COUNT

  The number of transactions committed since the last time the metrics were reset.

  This information is reset when the replica is restarted, or the user issues RESET REPLICA. It is not reset when the applier stops.

• TRANSACTIONS_ONGOING_COUNT

  The number of ongoing transactions. A transaction is marked as ongoing when its first event is scheduled on a worker; this counter decreases when the transaction is committed.

  This information is reset when the replica is restarted, or the user issues RESET REPLICA. It is not reset when the applier stops.

• TRANSACTIONS_PENDING_COUNT

  The number of pending transactions waiting to be applied. A transaction is considered to be pending when it is queued in the relay log, and is no longer considered pending when it is committed. This number includes any ongoing transactions.

  This information is reset when the replica is restarted, or the user issues RESET REPLICA. It is not reset when the applier stops.

  When the server is restarted with existing relay logs, the number of pending transactions is known only after all old relay logs have been consumed. Until this has taken place, the value shown in this column is NULL.

• TRANSACTIONS_COMMITTED_SIZE_BYTES_SUM

  The total size, in bytes, of all transactions committed since the last time the metrics were reset.

  This information is reset when the replica is restarted, or the user issues RESET REPLICA. It is not reset when the applier stops.

• TRANSACTIONS_ONGOING_FULL_SIZE_BYTES_SUM

  The total size (in bytes) of all transactions currently being executed by workers. This increases for the first GTID seen or (for compressed transactions) the first transaction payload event scheduled for each transaction, and decreases when a transaction is committed.

  This information is reset when the replica is restarted, or the user issues RESET REPLICA. It is not reset when the applier stops.

• TRANSACTIONS_ONGOING_PROGRESS_SIZE_BYTES_SUM

  The size (in bytes) of the parts of transactions that have already been executed for ongoing transactions. This increases for the first GTID seen or (for compressed transactions) the first transaction payload event scheduled for each transaction, and decreases when a transaction is committed.

  This information is reset when the replica is restarted, or the user issues RESET REPLICA. It is not reset when the applier stops.

• TRANSACTIONS_PENDING_SIZE_BYTES_SUM

  The size, in bytes, of all transactions awaiting execution. This increases for the first GTID seen or (for compressed transactions) the first transaction payload event scheduled for each transaction, and decreases when a transaction is committed.

  This information is reset when the replica is restarted, or the user issues RESET REPLICA. It is not reset when the applier stops.

  A transaction is considered to be pending when it is queued in the relay log, and is no longer considered pending when it is committed. This number includes any ongoing transactions.

  When the server is restarted with existing relay logs, the number of pending transactions is known only after all old relay logs have been consumed. Until this has taken place, the value shown in this column is NULL.

• EVENTS_COMMITTED_COUNT

  The number of events committed since the last time the metrics were reset. For compressed transactions, this counts embedded events but not payload events.

  This information is reset when the replica is restarted, or the user issues RESET REPLICA. It is not reset when the applier stops.

• WAITS_FOR_WORK_FROM_SOURCE_COUNT

  The number of times spent waiting for work from upstream, that is, waiting for the relay log to grow, since the last time the metrics were reset. It is possible for this counter to increase when there is no work to be done; the coordinator executes not one, but rather a continuous series of waiting cycles.

  This information is reset when the replica is restarted, or the user issues RESET REPLICA. It is not reset when the applier stops.

• WAITS_FOR_WORK_FROM_SOURCE_SUM_TIME

  The time spent waiting for work from upstream, that is, waiting for the relay log to grow since the last time the metrics were reset.

  This information is reset when the replica is restarted, or the user issues RESET REPLICA. It is not reset when the applier stops.

• WAITS_FOR_AVAILABLE_WORKER_COUNT

  The number of times the coordinator has waited while scheduling a transaction until a worker has become available.

  This information is reset when the replica is restarted, or the user issues RESET REPLICA. It is not reset when the applier stops.

• WAITS_FOR_AVAILABLE_WORKER_SUM_TIME

  The aggregate time in nanoseconds that the coordinator has waited while scheduling a transaction until a worker has become available.

  This information is reset when the replica is restarted, or the user issues RESET REPLICA. It is not reset when the applier stops.

• WAITS_COMMIT_SCHEDULE_DEPENDENCY_COUNT

  The number of times the coordinator has waited for a preceding or dependent transaction to commit.

  This information is reset when the replica is restarted, or the user issues RESET REPLICA. It is not reset when the applier stops.

• WAITS_COMMIT_SCHEDULE_DEPENDENCY_SUM_TIME

  The aggregate time that the coordinator has waited for a preceding or dependent transaction to commit.

  This information is reset when the replica is restarted, or the user issues RESET REPLICA. It is not reset when the applier stops.

• WAITS_FOR_WORKER_QUEUE_MEMORY_COUNT

  The number of times the coordinator has waited while scheduling a worker to process an event until the worker has reduced the queue size below replica_pending_jobs_size_max bytes.

  This information is reset when the replica is restarted, or the user issues RESET REPLICA. It is not reset when the applier stops.

• WAITS_FOR_WORKER_QUEUE_MEMORY_SUM_TIME

  The aggregate time that the coordinator has waited to schedule a worker to process an event until the worker has reduced the queue size below replica_pending_jobs_size_max bytes.

  This information is reset when the replica is restarted or the channel on the replica is deleted. It is not reset when the applier stops.

• WAITS_WORKER_QUEUES_FULL_COUNT

  The number of times the coordinator has waited because there were no empty slots for adding more tasks to the worker queue.

  This information is reset when the replica is restarted or the channel on the replica is deleted. It is not reset when the applier stops.

• WAITS_WORKER_QUEUES_FULL_SUM_TIME

  The aggregated time the coordinator has waited because there were no empty slots for adding more tasks to the worker queue.

  This information is reset when the replica is restarted or the channel on the replica is deleted. It is not reset when the applier stops.

• WAITS_DUE_TO_COMMIT_ORDER_COUNT

  The cumulative time spent by the coordinator on reading events from the relay log since the metrics were reset.

  This information is reset when the replica is restarted or the channel on the replica is deleted. It is not reset when the applier stops.

• WAITS_DUE_TO_COMMIT_ORDER_SUM_TIME

  The number of times workers have waited for preceding transactions to commit before they could commit their own transactions.

  This information is reset when the replica is restarted or the channel on the replica is deleted. It is not reset when the applier stops.

• TIME_TO_READ_FROM_RELAY_LOG_SUM_TIME

  The time in nanoseconds that workers have waited for preceding transactions to commit before they could commit their own trnsactions.

  This information is reset when the replica is restarted or the channel on the replica is deleted. It is not reset when the applier stops.

The replication_applier_metrics table was added in MySQL 9.1.0.