At the moment multiline logs and stacktraces generated in the CloudFoundry syslog as multilines instead of a single line entity. Streaming these logs then for example to Splunk each line of the stacktrace would be one line in SPLUNK.
That makes the search and readable of the logs very hard.
CloudFoundry defines directly something for Java Applications to write stacktraces as one entity to use a specific library from github.
App Logging in Cloud Foundry | Cloud Foundry Docs
The java multiline binaries would be these: loggregator-release/java-multi-line-work-around.md at main · cloudfoundry/loggregator-release (github.com)
Could you please change the output for multilines and stacktraces to be written in the syslog as one entity?
We resolved this topic for our own log statements and put them in a single line of json structure. And then it was totally readable in SPLUNK.
We funnel all our app logs into Open Search Dashboards and face the exact same issue. Would be nice if it was not split like this. When I spoke to Mendix we were told "Unfortunately, there is no out of the box solution available for this from Mendix."
I faced again this situation in troubleshooting and looking into the logs it is hard to read as the stacktrace of an exception is a single line.
The full stacktrace can be surrounded with different messages