Jenkins currently seems to log its stacktraces across multiple systemd journal messages - one journal message per stacktrace line. This can be seen when doing 

journalctl --output=json

for example, where each entry has a key-value-pair like this:

"MESSAGE":"\tat PluginClassLoader for workflow-job//org.jenkinsci.plugins.workflow.job.WorkflowRun.getExecution(WorkflowRun.java:719)"

This increases the complexity when trying to parse/suppress specific messages inside monitoring solutions as this does not allow for proper grouping/context-based handling.

For this reason, I would prefer a way to tell Jenkins not to split this across multiple messages.