HAProxy

Auto-Detection

Built-in Metrics

Prometheus Metrics

Logs

Bleemeo monitors HAProxy through its stats page (all versions) or its native Prometheus endpoint (2.0+).

Auto-Detection

Glouton automatically detects HAProxy instances running on the system.

Built-in Metrics

Built-in metrics are gathered from the HAProxy CSV stats page. You need to enable the stats page and tell Glouton where to find it.

Service Configuration

Add a statistics endpoint to your HAProxy configuration:

frontend api-http
    bind 0.0.0.0:80
    stats enable
    stats uri     /statistics

Glouton Configuration

Linux

sudo tee /etc/glouton/conf.d/99-haproxy.conf > /dev/null << 'EOF'
service:
  - type: "haproxy"
    address: "127.0.0.1"
    port: 80
    stats_url: "http://my-server/statistics"
    # For authenticated access, use:
    # stats_url: "http://username:password@my-server/statistics"
EOF

Windows

Create or edit the file C:\ProgramData\glouton\conf.d\99-haproxy.conf:

service:
  - type: "haproxy"
    address: "127.0.0.1"
    port: 80
    stats_url: "http://my-server/statistics"
    # For authenticated access, use:
    # stats_url: "http://username:password@my-server/statistics"

Glouton automatically detects configuration changes.

Metric	Description
service_status	Status of HAProxy
haproxy_act	Number of active servers
haproxy_bin	Network traffic received from clients in bytes per second
haproxy_bout	Network traffic sent to clients in bytes per second
haproxy_ctime	Average time spent opening a connection in seconds
haproxy_dreq	Number of requests denied by HAProxy per second
haproxy_dresp	Number of responses denied by HAProxy per second
haproxy_econ	Number of failed connections per second
haproxy_ereq	Number of requests in error per second
haproxy_eresp	Number of responses in error per second
haproxy_qcur	Number of currently queued requests
haproxy_qtime	Average time a request spent in queue in seconds
haproxy_req_tot	Number of HTTP requests per second
haproxy_rtime	Average response time in seconds
haproxy_scur	Number of sessions opened
haproxy_stot	Number of sessions per second
haproxy_ttime	Average session time in seconds

Prometheus Metrics

Since version 2.0, HAProxy exposes a native Prometheus endpoint that Glouton can scrape directly. This provides more detailed metrics than the stats page approach.

Service Configuration

Add a dedicated frontend to your HAProxy configuration (/etc/haproxy/haproxy.cfg):

frontend prometheus
    bind *:8405
    http-request use-service prometheus-exporter if { path /metrics }
    no log

Reload HAProxy to apply the change:

sudo systemctl reload haproxy

You can verify the endpoint is working by visiting http://localhost:8405/metrics.

Glouton Configuration

Linux

sudo tee /etc/glouton/conf.d/99-haproxy.conf > /dev/null << 'EOF'
metric:
  prometheus:
    targets:
      - url: "http://localhost:8405/metrics"
        name: "haproxy"
        allow_metrics:
          - "haproxy_process_current_connections"
          - "haproxy_process_uptime_seconds"
          - "haproxy_frontend_bytes_in_total"
          - "haproxy_frontend_bytes_out_total"
          - "haproxy_backend_http_responses_total"
          - "haproxy_server_status"
          - "haproxy_frontend_current_sessions"
          - "haproxy_backend_current_sessions"
          - "haproxy_server_current_sessions"
          - "haproxy_backend_connect_time_average_seconds"
          - "haproxy_backend_response_time_average_seconds"
EOF

Windows

Create or edit the file C:\ProgramData\glouton\conf.d\99-haproxy.conf:

metric:
  prometheus:
    targets:
      - url: "http://localhost:8405/metrics"
        name: "haproxy"
        allow_metrics:
          - "haproxy_process_current_connections"
          - "haproxy_process_uptime_seconds"
          - "haproxy_frontend_bytes_in_total"
          - "haproxy_frontend_bytes_out_total"
          - "haproxy_backend_http_responses_total"
          - "haproxy_server_status"
          - "haproxy_frontend_current_sessions"
          - "haproxy_backend_current_sessions"
          - "haproxy_server_current_sessions"
          - "haproxy_backend_connect_time_average_seconds"
          - "haproxy_backend_response_time_average_seconds"

Glouton automatically detects configuration changes.

Customize collected metrics

The metrics below are our recommended defaults. You can adjust the allow_metrics list to include any metric from HAProxy’s Prometheus documentation, or use deny_metrics to exclude specific ones. See Metrics Filtering for details.

Metric	Description
haproxy_process_current_connections	Current number of connections on the HAProxy process
haproxy_process_uptime_seconds	HAProxy process uptime in seconds
haproxy_frontend_bytes_in_total	Total bytes received by all frontends
haproxy_frontend_bytes_out_total	Total bytes sent by all frontends
haproxy_backend_http_responses_total	Total HTTP responses from backends, by status code
haproxy_server_status	Current status of each backend server (1 = UP, 0 = DOWN)
haproxy_frontend_current_sessions	Number of current sessions on each frontend
haproxy_backend_current_sessions	Number of current sessions on each backend
haproxy_backend_connect_time_average_seconds	Average backend connection time in seconds
haproxy_backend_response_time_average_seconds	Average backend response time in seconds

Logs

Glouton automatically collects HAProxy logs when available.

Monitoring Troubleshooting

See Troubleshoot a Service Check or Missing Metrics for what applies to every service: finding the address and port Glouton really uses, what each check message means, and how to read the collection error — which does not appear in the agent logs at the default level.

The Service Check is not OK

The check only opens the TCP port — it sends no command — so it fails when the port is wrong (Connection refused), when a firewall drops the packets (connection timed out after 10 seconds), or when the configured address cannot be parsed (Invalid TCP address). A green check proves the port is open, and nothing more.

Metrics are Missing

Glouton has no default for the HAProxy statistics endpoint: without stats_url in the service configuration, no collector is created at all and the service reports only service_status.

Reproduce what Glouton does, from the machine where the agent runs:

curl -s "http://127.0.0.1:8404/stats;csv" | head -2

What you see	Cause	Fix
nothing — the collector is never created	stats_url not set	Set stats_url to the stats page or to the admin socket
404, or a connection error	The stats page is not enabled	Add a stats enable / stats uri listener to haproxy.cfg
parsing errors	The URL does not return CSV	Append ;csv to the stats URL
401	The stats page requires authentication	Put the credentials in the URL: http://user:password@127.0.0.1:8404/stats;csv
permission denied	The admin socket is not readable by the agent	Give the glouton user access to the socket, or use the HTTP stats page