Getting Logstash _grokparsefailure though Grok Debugger throws no errors

Michael picture Michael · Mar 19, 2015 · Viewed 11.2k times · Source

I try to parse Check Point firewall Syslog logs with logstash and grok.

Example of a log entry:

<190>2015 Mar 19 12:40:55 fw1 <60031> User admin failed to login (wrong authentication) (Source IP:123.123.123.123 Via:HTTP)

I use this pattern:

<%{POSINT:syslog_pri}>%{YEAR} %{SYSLOGTIMESTAMP:syslog_timestamp} %{DATA:device} <%{POSINT:status}> User %{WORD:account} %{DATA:msg} (?:[(])%{DATA:msg1}(?:[)]) (?:[(])Source IP:%{IPV4:src} Via:%{WORD:protocol}(?:[)])

All fields are parsed well and show up in elasticsearch/kibana. The Grok Debugger works fine with this specific log/pattern combination. However, I keep receiving _grokparsefailure tags. Has anyone a hint how to get rid of them?

UPDATE: Here is my complete logstash configuration (most relevant part is the "Failed login" block):

input { 
    syslog {
        type => "syslog"
        port => 514
    }
}

filter {
    if [type] == "syslog" {
        geoip { source => "host" }  

        # Firewall rule fired
        if [message] =~ "packet" {
            grok {
                match => [ "message", "<%{POSINT:syslog_pri}>%{YEAR} %{SYSLOGTIMESTAMP:syslog_timestamp} %{DATA:device} <%{POSINT:status}> %{WORD:activity} %{DATA:inout} (?:[(])%{DATA:msg}(?:[)]) Src:%{IPV4:src} SPort:%{POSINT:sport} Dst:%{IPV4:dst} DPort:%{POSINT:dport} IPP:%{POSINT:ipp} Rule:%{INT:rule} Interface:%{WORD:iface}" ]
            }
        } 
        # Failed login
        else if [message] =~ "failed" {
            grok {
                match => [ "message", "<%{POSINT:syslog_pri}>%{YEAR} %{SYSLOGTIMESTAMP:syslog_timestamp} %{DATA:device} <%{POSINT:status}> User %{WORD:account} %{DATA:msg} (?:[(])%{DATA:msg1}(?:[)]) (?:[(])Source IP:%{IPV4:src} Via:%{WORD:protocol}(?:[)])" ]
            }
        }
        # Successful login/out
        else if [message] =~ "logged" {
            mutate {
                add_field => [ "userlogged", "%{host}" ]
            }
            grok {
                match => [ "message", "<%{POSINT:syslog_pri}>%{YEAR} %{SYSLOGTIMESTAMP:syslog_timestamp} %{DATA:device} <%{POSINT:status}> User %{DATA:account} %{WORD} %{WORD:action} (?:[(])Source IP:%{IPV4:src} Via:%{WORD:protocol}(?:[)])" ]
            }
        }
        else {
            grok {
                match => [ "message",  "<%{POSINT:syslog_pri}>%{SYSLOGTIMESTAMP:syslog_timestamp} %{SYSLOGHOST:syslog_hostname} %{DATA:syslog_program}(?:\[%{POSINT:syslog_pid}\])?: %{GREEDYDATA:syslog_message}" ]
            }
        }
    }   
}

output {
    elasticsearch {
        host => "localhost"
        protocol => "http"
    }
}

Answer

Michael picture Michael · Mar 19, 2015

It seems that the _grokparsefailure is thrown by the input plugin "syslog" which internally also uses grok. After replacing the input block with

input {
    tcp {
        port => 514
        type => syslog
    }
    udp {
        port => 514
        type => syslog
    }
}

I dind't receive any more the failure messages. This blog post helped me a lot.