Web Server Load Balancing with NGINX Plus

In a short five years, NGINX, Inc. has gone from powering zero to powering one in every six of the busiest web sites on the Internet. Datadog uses NGINX and chances are you do too. Datadog is pleased to announce the expansion of our current NGINX monitoring tool to support the additional features in NGINX Plus.

Launched in 2013, NGINX Plus is the fully supported, commercial version of NGINX that offers basic open source modules and additional enterprise‑grade features. In particular, this premium product provides:

Monitor NGINX Plus Metrics in Datadog

Datadog expands the existing plug‑in for NGINX Open Source to support the advanced live activity monitoring in NGINX Plus. Datadog now collects all metrics exposed by the NGINX Plus Status module, including:

  • Connections accepted, dropped, active, and idle
  • Bytes sent and received by each server zone
  • Number of health checks failed
  • Response counts, categorized by status code and upstream server

You can see a sample of all NGINX Plus metric names and their tagging structure here.

NGINX Connections and Upstreams

Tags are the linchpin of monitoring with Datadog. They are used everywhere to navigate data, define graphs, and create alerts easily, whether you have 5 or 5,000 servers. With tags you can slice and dice metrics, graphs, and alerts, and get an inventory of servers to quickly pinpoint the servers you wish to look into.

The new integration with NGINX Plus tags all NGINX Plus metrics for you with the correct upstream server and server zone, so that you can very easily see how traffic gets distributed between NGINX Plus servers, zones, and individual upstream servers. For example, in the graph below we are showing active connections per minute, split by upstream server.

Datadog graph showing number of active connections forwarded by NGINX Plus to each upstream server over a 1-hour period

Aggregate, Alert On, and Graph NGINX Plus Metrics

Datadog automatically creates a metric for every combination of upstream pool, upstream server, and server zone. These metrics can be aggregated in order to monitor an entire upstream pool or cluster of upstream pools. Furthermore, metrics from multiple NGINX Plus servers can be aggregated into a single, high‑level metric for the entire web service served via NGINX Plus.

You can graph any of these metrics on Datadog dashboards or receive alerts on them via email, PagerDuty, HipChat, Slack, or webhooks. Datadog’s high‑resolution historical data makes it a snap to conducting a drill‑down analysis for any event that happened over the last year.

Datadog graph showing an alert resulting from the number of 5xx exceeding a configured limit

Correlate Events from Other Systems with NGINX Plus Metrics

You can overlay events from the other 50+ Datadog integrations on top of the NGINX Plus metrics. With a side‑by‑side view of events and related dashboards of metrics, you can drastically cut the time you spend investigating and correlating events and metrics coming from different parts of your infrastructure.

A side-by-side view of NGINX requests by server zone and average system load

If you want the kind of visibility into your HTTP load balancing described in this article, first sign up for a free trial of NGINX Plus. Then try a 14‑day free trial of Datadog to instantly gain insight into NGINX metrics. Visit the Datadog NGINX Integration page to learn more about the Datadog NGINX integration.

Hero image
《NGINX 完全指南》2024 年最新完整版




F5, Inc. 是备受欢迎的开源软件 NGINX 背后的商业公司。我们为现代应用的开发和交付提供一整套技术。我们的联合解决方案弥合了 NetOps 和 DevOps 之间的横沟,提供从代码到用户的多云应用服务。访问 了解更多相关信息。