Qlik Replicate - safety mechanism for disk space

[Jon_Donker]

Hi All,

Hopefully a simple question.

Suppose "Task_1" and "Task_2" are on the same QR server.

Task_1's endpoint goes offline so QR starts back up changes first in memory and then on disk.

Will Task_1 keep backing up changes to the point where it forces Task_2 offline and potentially use all the disk on the server?

Or is there a safety mechanism to fail Task_1 when it approaches a disk space threshold?

Thanks.

[Arun_Arasu · Accepted solution]

Hi @Jon_Donker ,

Thank you for reaching out to the Qlik community!

Regarding Task_1's endpoint going offline, the task will automatically attempt to reconnect. After several retry attempts, if it remains unable to establish a connection, the task will terminate on its own.

As for the threshold, there is an option in Qlik Replicate under Resource Control that monitors memory and disk space usage. If these parameters exceed the defined threshold limits, the system will automatically close tasks. It's important to note that this setting operates at the server level, affecting all tasks, and cannot be applied to specific tasks.

For more detailed information, please refer to the user guide at the following link:

Qlik Replicate User Guide :https://help.qlik.com/en-US/replicate/May2023/Content/Global_Common/Content/SharedEMReplicate/Server...

If the provided above information helps please "Accept as Solution".

Regards

Arun

[Jon_Donker]

Perfect - thanks

[DesmondWOO]

Hi @Jon_Donker ,

Please allow me to provide further details regarding your questions. The outcome depends on your task settings and how Replicate handles the error.

Let's assume that the target endpoint is shut down. In this situation, Replicate may consider it a network error, which is a recoverable error. If the maximum retry count is set to -1 (infinite), Replicate will continuously attempt to establish a connection with the target database while offloading changes from memory to disk. It won't impact other running tasks.

However, if the disk space becomes full, it can lead to unexpected results as only partial information will be written to the disk and Replicate's repository. You may need to restore Replicate as the repository is damaged. If you have set up Resource Control for disk space, all tasks will be halted once the Critical Disk Space Utilization Threshold is reached. 

Regards,
Desmond

[Dana_Baldwin]

Hi @Jon_Donker 

Related to what @DesmondWOO shared, if your server gets to the critical disk threshold and stops the tasks, you must free enough space so that it is below the amount defined in the "high" threshold before tasks will resume.

Hope this helps!

Dana

[Jon_Donker]

Thanks all.

We have disk space alerting on the server so we will be alerted well before things get dire.

The system architects were  seeking some worse case scenario answers for regulatory compliance.