no metrics but all nginx log parameters appear

[tek0101]

Hi Martin!
I don't see any metrics except 0 errors when i make request to /metrics url but all parameters present in my nginx config and in exporter

My configs:

nginx-config :

	log_format main  '$remote_addr - $remote_user [$time_local] '
        '"$request" $status $body_bytes_sent "$http_user_agent" '
	'$request_time "$upstream_connect_time" "$upstream_header_time" 
       "$upstream_response_time"';

log exporter config:

<code>listen {
  port = 4040
  address = "0.0.0.0"
}

namespace "app" {
  format = "$remote_addr - $remote_user [$time_local] \"$request\" $status $body_bytes_sent  \"$http_user_agent\" $request_time \"$upstream_connect_time\" \"$upstream_header_time\" \"$upstream_response_time\""
 
  
  source {
    files = ["/var/log/nginx/access.log"]
  }
  labels {
    app = "app"
    environment = "test"
  }
}

exporter-log :

loading configuration file /etc/prometheus-nginxlog-exporter.hcl
No globs for /var/log/nginx/access.log
using configuration {Listen:{Port:4040 Address:0.0.0.0 MetricsEndpoint:/metrics} Consul:{Enable:false Address: Datacenter: Scheme: Token: Service:{ID: Name: Address: Tags:[]}} Namespaces:[{Name:app NamespaceLabelName: NamespaceLabels:map[] MetricsOverride:<nil> NamespacePrefix: SourceFiles:[/var/log/nginx/access.log] SourceData:{Files:[/var/log/nginx/access.log] Syslog:<nil>} Parser: Format:$remote_addr - $remote_user [$time_local] "$request" $status $body_bytes_sent  "$http_user_agent" $request_time "$upstream_connect_time" "$upstream_header_time" "$upstream_response_time" Labels:map[app:app environment:production] RelabelConfigs:[] HistogramBuckets:[] PrintLog:false OrderedLabelNames:[] OrderedLabelValues:[]}] EnableExperimentalFeatures:false EnableExperimentalFeaturesOld:false}
starting listener for namespace app
running HTTP server on address 0.0.0.0:4040, serving metrics at /metrics
2023/01/16 08:09:17 Waiting for /var/log/nginx/access.log to appear...

log-exporter-output:

curl http://localhost:4040/metrics
# HELP app_parse_errors_total Total number of log file lines that could not be parsed
# TYPE app_parse_errors_total counter
app_parse_errors_total 0
# HELP prometheus_nginxlog_exporter_build_info A metric with a constant '1' value labeled by version, revision, branch, and goversion from which prometheus_nginxlog_exporter was built.
# TYPE prometheus_nginxlog_exporter_build_info gauge
prometheus_nginxlog_exporter_build_info{branch="v1.10.0",goversion="go1.18.6",revision="9274be256a47378a9930164023d16e1958e96292",version="1.10.0"} 1

[martin-helmich]

At a quick glance: In your exporter config there are two spaces before \"$http_user_agent\", instead of one (like in the NGINX config); this will most likely cause the generated regex to not match any of the log lines, because the exporter will then also expect double spaces in your log lines.

Hope that helps.