In this article I'm going to discuss a couple of possible causes for failing Scheduled ETL jobs.
The most common cause is that there was not enough memory allotted to the software, more specifically Java to complete the job.
What happens in this case is there is a java heap space error and it crashes Tomcat/iDashboards service, which is usually set to automatically restart so other jobs may continue to run.
What to look for...
Check the idbdata.log files. These can be found either through the datahub System Tab > Logs or on the server in the ivizgroup folder C:\Program Files\iDashboards_SRV\ivizgroup\logs by default.
Look for this message "java.lang.OutOfMemoryError: Java heap space" it will likely be nearby the name of the failed ETL job.
What to do about it...
On the server open the Java options console by navigating to C:\Program Files\iDashboards_SRV\Server\bin and double clicking iDashboardsW.exe.
If you can't open the console this way you can do it via command line. Navigate to to the same bin folder and run command: idashboardsW.exe //ES//iDashboardsSRV
When the console loads click on the Java tab.
In the Java Options window scroll down until you see the -Xms/-Xmx option.
-Xms determines the minimal RAM that will be used. Nothing needs to be done here.
Set the -Xmx amount to suit your ETL usage. Alternatively you can remove the -Xmx line. This will allow access to all available memory.
After you apply the change, restart the iDashboards Service for the change to take effect.
If you still receive the heap space error after increasing or removing -Xmx you will likely need to increase the overall memory on the server.
Possible Cause #2
This cause was reported by a customer. They found that scheduled ETL jobs we're failing whenever there we're pending Windows Updates on the server.
We have not verified this correlation, however there is a lot of information online of people reporting pending updates causing a variety of issues from affecting software to performance issues.
If you find this is the reason your Scheduled ETL jobs are failing. You can turn off automatic updates and check for and run updates manually.