Oracle DataGuard. Endpoint reconnection for a Logstream task and associated tasks
In reference to: community.qlik.com/t5/Official-Support-Articles/Qlik-Replicate-Oracle-Source-Task-hangs-without-failures-or/ta-p/1958228
I have a powershell script running on the QLIK host that checks for task failures. I see this parameter referenced in that document:
Replicate preparation
The following procedure describes how to configure Replicate to handle a failover operation in a Data Guard environment:
Create a task that points to the primary database.
In the advanced settings for the primary Oracle source endpoint, set the supportResetLog=true internal parameter.
Run the task. It will be active until a failover occurs.
So, I've updated that in my QA environment.
The document goes on to say that I need to restart the failed task using the saved checkpoint for a failover.
Note, that after setting "supportResetLog=true"
I've found that the logstream task reconnects to the database following a failover and the associated sub tasks that use the logstream data keep running and don't lose any data.
Seeing this in my QLIK logs for the logstream task:
ORA-03113: end-of-file on communication channel Process
Then:
Trying to reconnect to the source database
Then it figures out the correct log to open as it starts with log 1 on the new primary database after the failover.
The LS task reconnects starting at the correct checkpoint and everything flows as intended.
Question: With the logstream task reconnecting automatically. Do I even need to restart the task with a checkpoint for logstream?
Note, I tried to restart using the logstream task that way. But then my associated subtasks all failed. I had to manually restart them all using the Advanced Run Options-> Tables are already loaded. Start processing changes from: -> Source change position (e.g. SCN or LSN): 0
When you say 'LS task" I assume you are refering to the LS parent task which is the task that is responsible for reading the CDC events from the redo log (please confirm). In order to better understand your sequence of events, please open a Salesforce case with the description above and attache to it the task diagnostic package of both log stream parent task and logs stream child task so we can see what was going on from task perspective and provide a better solution to your problem.
