Syslog input going to "%{[@metadata][beat]}-2019.05.28"

[tedder (ted)]

I don't know what information is needed, but I've started sending syslog input into logstash (and therefore into ELK).

My input simply specifies a port, the filter contains the following:

filter {
  if [type] == "syslog" {
    grok {
      match => { "message" => "%{SYSLOGTIMESTAMP:syslog_timestamp} %{SYSLOGHOST:syslog_hostname} %{DATA:syslog_program}(?:\[%{POSINT:syslog_pid}\])?: %{GREEDYDATA:syslog_message}" }
      add_field => [ "received_at", "%{@timestamp}" ]
      add_field => [ "received_from", "%{host}" ]
    }
    syslog_pri {
    }
    date {
      match => [ "syslog_timestamp", "MMM  d HH:mm:ss", "MMM dd HH:mm:ss" ]
    }
  }
}

I then added the following to syslog_pri:

    syslog_pri {
      add_field => { "[@metadata][type]" => "syslog" }
      add_field => { "[@metadata][beat]" => "syslog" }

Here's what the output looks like:

output {
  elasticsearch {
    hosts => ["localhost"]
    manage_template => false
    index => "%{[@metadata][beat]}-%{+YYYY.MM.dd}"
    http_compression => true
    sniffing => true
    sniffing_delay => 20
  }
}

Neither of those syslog_pri options seem to have helped. It doesn't have any _grokparsefailure, but it has syslog-ish fields like facility and severity, so I assume it is being picked up and being filtered, it's just that the output index is literally the string "%{[@metadata][beat]}-2019.05.28".

I left the date in to make it clear I'm not scrubbing it in any way- that's literally the index name.