Skip to main content
Announcements
UPGRADE ADVISORY for Qlik Replicate 2024.5: Read More
cancel
Showing results for 
Search instead for 
Did you mean: 
andile
Contributor II
Contributor II

Resuming a Full Load + CDC Task that has Segments

Hi
I have a task that runs both full load and cdc. I added segments because the table is huge. I 'Start Processing' and the full load process initiates. Once there's data in the target database, I stop the task. I then 'Resume Processing'. I noticed that it 'Truncates' the target table and starts from scratch.

Is this how it should behave. Shouldn't it 'Resume' from where it left off? Even the transferred count resets to 0. Basically a fresh full load. My concern is that should I continue using segments and then the task in future fails, after fixing the issue and Resuming Processing, the task with start from scratch and do a fresh full load as I have mentioned this behavior above.

Please assist and clarify.

Labels (1)
2 Solutions

Accepted Solutions
SwathiPulagam
Support
Support

Hi @andile ,

 

Seems like you are stopping the task before the Full Load is finished. During the Full load process, if you stop the task and Resume means again the Full load will initiate, that is the expected behavior.

You have to wait until the Full Load completes to stop the task. 

 

Thanks,

Swathi

View solution in original post

Dana_Baldwin
Support
Support

Hi @andile 

Fully agree with Swathi. Tasks can only resume from where they left off when they are running in change processing mode (and of course source redo/transaction logs are still available).

Thanks,

Dana

View solution in original post

3 Replies
SwathiPulagam
Support
Support

Hi @andile ,

 

Seems like you are stopping the task before the Full Load is finished. During the Full load process, if you stop the task and Resume means again the Full load will initiate, that is the expected behavior.

You have to wait until the Full Load completes to stop the task. 

 

Thanks,

Swathi

Dana_Baldwin
Support
Support

Hi @andile 

Fully agree with Swathi. Tasks can only resume from where they left off when they are running in change processing mode (and of course source redo/transaction logs are still available).

Thanks,

Dana

andile
Contributor II
Contributor II
Author

Thanks, it clear. I have a huge table that runs full-load for days. When it fails, while still on full-load - say 90%, I have to try again from the beginning. I thought using segments were going to solve that problem. The full load has been separated into subsets using the dates so that when one subset fails, it doesn't affect the data that already been loaded.