Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
Feb 18, 2022 8:25:23 AM
Feb 16, 2022 7:24:47 PM
The current behavior of stopping task is as follows: the process will wait for open transaction completion. If they are not completed until timeout of a few minutes - the task will be killed
Replicate will wait for 30 minutes after stopping the task before the timeout happens.
Reducing the size of the bulk may help in some cases as it will get less chances of getting the task killed but is not always the case.
As for the "stop timeout occurred" error, when it happens there maybe times when there will be a leftover or ghost process for the task.
When this happens and you try to resume or start the task, it will look like that the task is starting but with no progress.
To resolve this type of issue, what you can do is to stop the task and check for the repctl processes. There should be no repctl process running for a stopped task. If there is one then you need to kill the repctl process.
Hello,
We need to decrease the stop timeout parameter which is set 30 minutes default.
Is there any repctl.cfg parameter or something else to configure this parameter?
Many Thanks.
Hey @kutay_cilingiroglu ,
The timeouts after 30 minutes is by design for Replicate. When you stop a task, it will wait for 30 minutes before it gets a timeout error.
If you need an enhancement request regarding the stop timeout, please visit our Ideations page. As a Qlik Community member, you can actively engage with our Product Management team, vote on a product idea coming for other Qlik customers, submit your own ideas and get feedback from other members.
https://community.qlik.com/t5/Suggest-an-Idea/idb-p/qlik-ideas
To get started please see our article: "Getting Started with Ideas": https://community.qlik.com/t5/Qlik-Product-Insight-Blog/Getting-Started-with-Ideas/ba-p/1684011
Please let us know if you have any additional questions!
Best,
Kelly
In case you are not aware, when a task times out while stopping, best practice is to start the task from timestamp taking any latency the task had into consideration, rather than resuming the task.
Regards,
Dana