Struggling to debug existing ELK installation with fluent

I was handled an existing implementation of ELK on hosts running in AWS. The web server instances are collecting logs from many sources but what’s most critical is nginx access files (we are multi-tenant and each domain has its own log file) which are secure forwarded to a log server running fluentd and EK as its index source and front end. Problem is these servers are collecting hundreds of requests every minute and I don’t know where to add debug without being flooded with output and having to sift through it.
A simple example might be how do I see these records before and after:

# Reformat NGINX log records into ELK format
<filter nginx.access.**>
  @type parser
  key_name message
  reserve_data yes
    @type regexp
    expression /^(?<remote>[^ ]*) (?<host>[^ ]*) (?<user>[^ ]*) \[(?<time>[^\]]*)\] "(?<method>\S+)(?: +(?<path>[^\"]*) +\S*)?" (?<code>[^ ]*) (?<size>[^ ]*) (?<duration>[^ ]*)(?: "(?<referer>[^\"]*)" "(?<agent>[^\"]*)" "(?<forwardedfor>[^\"]*)")?(?:\s+"(?<X-Station-Id>\S+)")?(?:\s+"(?<X-Cashdrawer-Id>\S+)")?$/
    time_format %d/%b/%Y:%H:%M:%S.%L %z
    time_key time
    types duration:float
# Change duration to number of milliseconds not seconds
<filter nginx.access.**>
  @type record_transformer
  enable_ruby true
    duration ${record['duration'] * 1000}

Do I simply add a $type stdout before and after? Is there a way to be more granular?