Skip to main content

JMX Metrics

All oracdc source connectors publish metrics about their activity that can be monitored through JMX.

MBean Names

ConnectorMBean
OraCdcRedoMinerConnectorsolutions.a2.oracdc:type=RedoMiner-metrics,name=<connectorName>,database=$ORACLE_SID_hostname
OraCdcLogMinerConnectorsolutions.a2.oracdc:type=LogMiner-metrics,name=<connectorName>,database=$ORACLE_SID_hostname

Available Metrics

Table and Partition Metrics

AttributeTypeDescription
TablesInProcessingCountintNumber of tables in processing
TablesInProcessingString[]Names of tables in processing
PartitionsInProcessingCountintNumber of partitions and subpartitions in processing
TableOutOfScopeCountintNumber of tables from the journal that do not meet the a2.include criteria

Redo Log Processing

AttributeTypeDescription
Last100ProcessedRedoLogsString[]Last 100 processed redo logs
LastProcessedRedoLogStringLast processed redo log
LastProcessedScnlongLast processed SCN
LastProcessedSequencelongLast processed log sequence
LastProcessedRedoLogTimeStringTime when the last redo log was processed
CurrentlyProcessedRedoLogStringRedo log currently being processed
CurrentFirstScnlongmin(FIRST_CHANGE#) for redo logs currently being processed
CurrentNextScnlongmax(NEXT_CHANGE#) for redo logs currently being processed
ProcessedRedoLogsCountintNumber of processed redo logs
ProcessedRedoLogsSizeGbfloatSize (GB) of processed redo logs

Timing and Performance

AttributeTypeDescription
StartTimeStringConnector start date and time (ISO format)
StartScnlongConnector start SCN
ElapsedTimeMillislongElapsed time in milliseconds
ElapsedTimeStringElapsed time in Days/Hours/Minutes/Seconds
ParseElapsedMillislongTime spent parsing raw bytes, milliseconds
ParseElapsedStringTime spent parsing, Days/Hours/Minutes/Seconds
ParsePerSecondintAverage number of records parsed per second
RedoReadElapsedMillislongTime spent reading redo logs, milliseconds
RedoReadElapsedStringTime spent reading redo logs, Days/Hours/Minutes/Seconds
RedoReadMbPerSecondfloatAverage MB per second of redo log reading

Record and Transaction Counts

AttributeTypeDescription
TotalRecordsCountlongTotal number of records processed
RolledBackRecordsCountlongNumber of rolled back records processed
RolledBackTransactionsCountintNumber of rolled back transactions processed
CommittedRecordsCountlongNumber of committed records processed
CommittedTransactionsCountintNumber of committed transactions processed
SentRecordsCountlongNumber of records already sent to Kafka broker
SentBatchesCountintNumber of Connector poll() calls

Lag Monitoring

AttributeTypeDescription
ActualLagSecondsintActual lag in seconds (SYSDATE - V$ARCHIVED_LOG.FIRST_TIME)
ActualLagTextStringActual lag in Days/Hours/Minutes/Seconds

DDL Processing

AttributeTypeDescription
DdlColumnsCountintNumber of successfully changed columns by DDL commands
DdlElapsedMillislongTime spent processing DDL commands, milliseconds
DdlElapsedStringTime spent processing DDL commands

Transaction Queue Metrics

AttributeTypeDescription
MaxTransactionSizeByteslongMaximum transaction size (off-heap storage), bytes
MaxTransactionSizeMiBfloatMaximum transaction size, MiB
MaxNumberOfTransInSendQueueintMaximum committed transactions in send queue
CurrentNumberOfTransInSendQueueintCurrent committed transactions in send queue
MaxNumberOfTransInProcessingQueueintMaximum committed transactions in processing queue
CurrentNumberOfTransInProcessingQueueintCurrent committed transactions in processing queue

Off-Heap Storage

AttributeTypeDescription
NumBytesWrittenUsingOffHeapMemlongTotal bytes written to Chronicle Queue memory mapped files
GiBWrittenUsingOffHeapMemfloatTotal GiB written to Chronicle Queue memory mapped files

Capacity Planning

Use the following metrics to determine required disk space for Chronicle Queue (off-heap storage), Docker tmpfs mounts, or Kubernetes emptyDir volumes:

  • MaxTransactionSizeMiB - peak memory for a single transaction
  • MaxNumberOfTransInProcessingQueue - peak concurrent transactions
  • GiBWrittenUsingOffHeapMem - total off-heap data written over time