The replication stream experiences a prolonged period of no data activity.
To help find the exact cause of your Extract stopping, could you share the listed right before this message? If you noticed any recent database changes or high system workloads , sharing those details would also help narrow it down. Share public link
However, when you see this message immediately after an ABEND or ERROR , it serves as a tombstone for a failed process. In those cases, it is a diagnostic marker indicating that the database mining component was forced to shut down. By understanding the distinction between a routine stop and an abort caused by missing logs or connectivity issues, DBAs can drastically reduce recovery time and maintain a healthy, continuous replication environment. Always check the database alert log in conjunction with the OGG report file to determine the true story behind the detach.
This guide explains what this message means, why it occurs, and how to troubleshoot it if it accompanies an unexpected system halt. Understanding the Architecture
Essentially, the database is acknowledging that the GoldenGate process has stopped requesting data and the session has been closed cleanly. Common Scenarios Where This Occurs The replication stream experiences a prolonged period of
: If accompanied by an error (like OGG-01668 ), it indicates the process stopped unexpectedly due to a missing trail file or privilege issue. Troubleshooting Steps
The most common cause is an administrative command. Issuing in GGSCI will gracefully detach the client from the Log Mining Server before stopping it. A successful detach ensures that no data is lost and that the Extract's checkpoint is written cleanly.
Drop the capture using DBMS_CAPTURE_ADM.DROP_CAPTURE . Register Again: GGSCI> REGISTER EXTRACT DATABASE Start Extract: GGSCI> START EXTRACT Step 4: Check for Resource Constraints
The Oracle GoldenGate (OGG) capture client has been successfully detached from the GoldenGate capture process. Share public link However, when you see this
When the GoldenGate Manager process ( MGR ) is restarted, it can temporarily pause and then resume connections. During this controlled restart, Extract processes may gracefully detach. Similarly, planned database maintenance may involve the Log Mining Server pausing, causing the client to detach intentionally.
Issuing a STOP EXTRACT command via GGSCI or Admin Client cleanly closes the OS process, triggering a graceful detachment from the database mining queue.
Set up alerts for high CPU usage and low disk space in the FRA.
The successful detachment of the OGG capture client from the GoldenGate capture process implies that: Always check the database alert log in conjunction
This implies the detachment was involuntary. Something—like a database crash, a memory exhaustion issue (SGA/Streams Pool), or a permissions change—forced the client off. Troubleshooting Forced Detachments
If the detachment occurs unexpectedly or is coupled with errors, use the following matrix to identify and remediate the underlying issue:
Are there any in the database alert log around the same timestamp?
Look for trailing errors, memory allocations, or specific ORA- errors right before the process terminated. Step 2: Cross-Reference the Database Alert Log
Ensure the STREAMS_POOL_SIZE has a minimum fixed allocation (at least 1GB to 2GB per Integrated Extract) rather than relying entirely on Automatic Memory Management.