How to configure Telegraf Logparser to parse stack trace

We are moving over to InfluxDB from ElasticSearch. We used logstash for log aggregation and I need to do something similar using InfluxDB. It looked like Telegraf and Logparser are a good replacement. Logstash provided a nice way of specifying the stack trace with codecs and multilines. I read that LogParser borrows quite a bit from Logstash and it too has an idea of multine, but it is not clear from the docs how to get it to work.

We are using log4net with the following layout (quite standard)
%date [%thread] %-5level %logger - %message%newline"

The %message field contains the stack trace too when using the Log.Error(msg, exception) overload.

Currently the logparser plugin cannot handle parsing multiple line messages. We have an issue open for adding support if you would like to subscribe to notifications.

I am not sure if the Multline feature is already built into the telegraph.
I want to parse JBOSS logs and it works for a one-line log:

Log: 2019-12-12 10:40:48,487 INFO [] (EJB default - 9) Aktuellster Vorlagenstand im Cache. Kein Cache-Refresh nötig.

measurement = “kiss24_log”
TimeZone = “Europe/Berlin”

patterns = [’^%{TIMESTAMP_ISO8601:timestamp:ts-“2006-01-02 15:04:05.000”} %{GREEDYDATA:Level} [%{GREEDYDATA:Logger}] (%{GREEDYDATA:Thread}) %{GREEDYDATA:Message}$’]

Unfortunately, I don’t find any instructions for Multiline. Can someone help.

2019-12-17 04:54:15,898 INFO [] (MSC service thread 1-7) WFLYEJB0473: JNDI bindings for session bean named ‘BriefinhaltService’ in deployment unit ‘deployment “beipackverwaltung.war”’ are as follows: