Qlik Replicate and Oracle source endpoint: consistently a 30-minute latency

[john_wang]

When working with the Oracle source endpoint, there is consistently a 30-minute latency. After an update is performed on the source, it takes 30 minutes for the changes to be reflected in the target Kafka. No configuration in the task explicitly sets this delay.

Environment:

• Qlik Replicate All versions
• Oracle All Versions

Resolution

To address this issue, consider one of the following options:

• Enable Qlik Replicate to access online redo logs.
  
  
• Shorten the automatic archive interval in Oracle, the frequency of redo log archival depends on how quickly the online redo log files are filled up.
  
  Oracle does not archive redo logs based on time intervals but rather based on log switches, which occur when an online redo log file becomes full and the database starts writing to the next redo log file.
  
  
• Manually force the changes to be written from the online redo logs to the archived redo log files by executing the following SQL statement:
  
  ALTER SYSTEM SWITCH LOGFILE;

Cause

Qlik Replicate captures changes from archived redo logs only. If changes on the source are still stored in Online Redo Logs during a given period, Qlik Replicate cannot retrieve them from the archived redo logs until approximately 30 minutes have passed.

In the Oracle source endpoint advanced property, Use archived redo logs only is ticked:

The task log reads:

2024-08-18T11:22:33 [SOURCE_CAPTURE ]I: Oracle CDC will access Archived Redo logs only

Internal Investigation ID(s)

00293052