Set up OpenTelemetry: 1-hour live workshop

Overview

You can now stream your Uptrends synthetic monitoring data into any observability platform or data pipeline that supports OpenTelemetry. That includes tools such as Grafana, Elastic, Datadog, ITRS Analytics, and any other platform that can receive OpenTelemetry data.

Each synthetic check in Uptrends generates a unique Correlation ID that your services can read and propagate across all tiers of your application. This enables you to correlate Uptrends checks with your existing metrics, logs, traces, and events for complete end-to-end visibility.

Who is it for?

This workshop is designed for SREs, DevOps engineers, platform engineers, and developers who want to:

1. Bring different signals together on one screen, for example:

   a. Synthetic measurements from Uptrends

   b. Infrastructure metrics

   c. Backend logs and traces

2. Follow a request made by an Uptrends monitor through the distributed backend, using a single Correlation ID per monitor check to join everything.

Example use case

Consider a scenario where your sign-up page experiences performance issues:

An Uptrends transaction monitor detects that the sign-up page fails to load within 30 seconds and triggers an alert. At the same time, Uptrends sends this check result into your observability platform using OpenTelemetry.

The operator who receives the alert opens the service dashboard in your observability tool. In the response time graph, they click the most recent data point to open the trace for this check. The trace combines data from the Uptrends transaction and from other systems that use the same Correlation ID. The trace reveals that a database call from the sign-up service timed out after 30 seconds.

With the root cause identified, you can take immediate corrective action.

This approach combines the external user perspective from Uptrends with the internal system view from your observability platform, providing comprehensive insight into the issue.

How does it work in practice?

Our team guides you through the complete setup in a hands-on, collaborative workshop session.

What we do together

In a 1-hour session with our product experts:

1. Connect your endpoint

   a. You provide a publicly accessible OpenTelemetry OTLP gRPC endpoint, secured with an API key or a username and password.

   b. If you prefer HTTP, we can discuss that as well.

2. Configure the Uptrends Exporter

   a. We configure the exporter in Uptrends while you are on the call.

   b. We send a few test checks and verify that you can see Uptrends data in your observability platform.

3. Check distributed tracing and correlation

   a. We confirm that traces look right and that attributes like Correlation ID are visible and usable.

If you like, we can also give you a short demo during this session and share a link to our OpenTelemetry webinar on demand for your wider team.

A few technical notes

Metric to trace navigation

If you want to click from a metric data point straight to the matching trace, your observability platform needs to support OpenTelemetry exemplars. Not every OpenTelemetry backend does this yet. If exemplars are not supported, you still get all the traces, you just navigate by time window and attributes instead of a single click.

Distributed tracing with Correlation ID

For full end-to-end tracing, your software needs to accept and pass on the Correlation ID that Uptrends sends in the X-Correlation-ID HTTP header. Your services can then include this ID in their own spans, making it easy to join everything.

Topics we will also discuss in the workshop

To make the exporter work well in your organization, we will ask a few questions.

Organization and grouping of data

Many OpenTelemetry-based systems have limited fine-grained access control. We want to understand how you organize your data:

• Do you separate telemetry by team or service in your platform today?
• Do you want all Uptrends monitors to send data to one endpoint, or different endpoints per team?
• Uptrends can send to different endpoints at monitor group level. Would that fit your setup?
• Would you consider running a single OpenTelemetry collector and doing routing and data changes there?

Security

We will check what you need to feel comfortable from a security point of view:

• What are your security requirements for the OpenTelemetry endpoint?
• Is protecting the endpoint with an API key or username and password enough?
• Is your endpoint publicly reachable?
• Do you want IP allow listing on your OpenTelemetry collector?

Resiliency

We also want to understand your expectations if your endpoint is not reachable:

• How long should Uptrends keep data that cannot be sent, and how often should we retry?
• Does your platform have limits on how old incoming data can be?
• Do you need rate limits on how fast we send data?

What does Uptrends need from you

Before or during the workshop, it helps if you can provide:

1. The URL endpoint of your observability platform or OpenTelemetry collector.

2. The authentication details needed to send data:

   a. For basic authentication: a username and password, which we store securely in our vault.

   b. For other methods: an API key or secret, and the header name or gRPC metadata key we should use.

We use a secure vault to store these details. You only need to send us the vault item name or ID if you already use our vault items internally.

(Optional) OpenTelemetry attribute settings

Depending on how you organize services in your observability platform, you may want to adjust a few OpenTelemetry attributes. The default settings are:

If you’re unfamiliar with these attributes, don’t worry, different platforms display and use them differently. During the workshop, we’ll review these settings with you and tailor them to your specific environment.