Logstash elasticsearch output plugin crashes elasticsearch

[hagit]

Hi,

I am using logstash 6.5.4 with elasticsearch output plugin to index 15M documents and it keeps crashing my elasticsearch (6.5.4). I'm getting:

[WARN ][logstash.outputs.elasticsearch] Marking url as dead. Last error: [LogStash::Outputs::ElasticSearch::HttpClient::Pool::HostUnreachableError] Elasticsearch Unreachable: [http://localhost:9200/][Manticore::SocketTimeout] Read timed out {:url=>http://localhost:9200/, :error_message=>"Elasticsearch Unreachable: [http://localhost:9200/][Manticore::SocketTimeout] Read timed out", :error_class=>"LogStash::Outputs::ElasticSearch::HttpClient::Pool::HostUnreachableError"}
[ERROR][logstash.outputs.elasticsearch] Attempted to send a bulk request to elasticsearch' but Elasticsearch appears to be unreachable or down! {:error_message=>"Elasticsearch Unreachable: [http://localhost:9200/][Manticore::SocketTimeout] Read timed out", :class=>"LogStash::Outputs::ElasticSearch::HttpClient::Pool::HostUnreachableError", :will_retry_in_seconds=>2}

My configuration:

input {  
    jdbc { 
        jdbc_connection_string => "jdbc:sqlserver://localhost;database=XXX;user=xxx;password=xxx"
        jdbc_user => nil
        jdbc_driver_library => "C:\Program Files\Microsoft JDBC Driver 6.2 for SQL Server\sqljdbc_6.2\enu\jre8\sqljdbc4-3.0.jar"
        jdbc_driver_class => "com.microsoft.sqlserver.jdbc.SQLServerDriver"

        jdbc_fetch_size => 500

       statement => "..."
    }
}

filter {
    ruby {
        code => 'event.set("[my_suggest][contexts][location]", [{"lat" => event.get("my_suggest.contexts.location.lat"),"lon" => event.get("my_suggest.contexts.location.lon")}]) if event.get("my_suggest.contexts.location.lon");'
    }
    
    mutate {
        remove_field => "my_suggest.contexts.location.lat"
        remove_field => "my_suggest.contexts.location.lon"
        rename => {
            "my_suggest.input" => "[my_suggest][input]"
            "my_suggest.weight" => "[my_suggest][weight]"
        }
    }
}

output {
    elasticsearch {
        hosts => ["localhost:9200"]
        index => "my_index"
    }  
}

Please advise. Thank you.

[Christian_Dahlqvist]

Is there anything in the Elasticsearch logs??

[hagit]

Thanks for the fast response.
In elasticsearch_production.log I get this warning:

[WARN ][o.e.m.j.JvmGcMonitorService] [node_master_data_1] [gc][7165] overhead, spent [608ms] collecting in the last [1s]

And indeed increasing the heap size in jvm.options seems to solve the problem.
Sorry - I missed this log between the other files under the directory...