Also, get rid of the leading '%{GREEDYDATA:nonsense}', it will make the pattern much more expensive if a line does not match. Starting the pattern with the : that follows that limits the number of places in the log line where it has to starting trying to match.
I would actually dissect that instead of using grok
If you do not want some fields you can replace %{pri} with %{} etc. -- it will still consume the text between the delimiters but not store it as a field.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.