I tried the grok. It seems working but if there are multiple rows between started and ended words, Logstash creates for each row a different document in the index.
I need the text between these two words to be in the one document.
Maybe the setting in input plugin are incorrect?