Skip to main content

Quick Start

Create one application, install an agent, confirm the service, open a trace, and save one policy.

Prerequisites​

  • Your account can onboard and view APM in the target organization.
  • Node Management already has a region. The agent uses that region's internal endpoint only.
  • The target host can reach the in-product agent download and the selected endpoint.
  • The .NET agent currently covers Linux x86_64 glibc.

Create an application and install​

  1. Open Integration → Applications and create an application with a name and region.
  2. Open the install instructions and choose Java, Python, Node.js, Go, or .NET.
  3. Optional: set a sample rate on this install script. It is written only into this script, defaults to full sampling, is not saved on the application, and does not change agents that are already running.
  4. Follow the offline steps on the target host. Do not switch the instructions to a public token.

UI Guide:

Screenshot to be added.

  • What to check: The region and language match the host. Keep the download URL and endpoint exactly as the page shows them.

Confirm the service directory​

  1. After data arrives, open Services → Services.
  2. Confirm the service is in the current organization. The active window is about 15 minutes.
  3. Archive a service you no longer watch, and restore it later if needed.
  4. Open Integration → Reporting instances and confirm the instance. A superuser can filter unassigned objects and set an organization. The normal list does not mix unassigned objects into the current organization.

Inspect topology and traces​

  1. Open Services → Service topology. Keep the window within 7 days. Downstream dependencies without an agent can appear here, but they are not added to the service directory.
  2. Open Explore → Traces. Filter by time, service, and environment. The longest query window is 7 days. The choices are 15m, 1h, 4h, 1d, and 7d.
  3. Open one trace and use the waterfall, flame graph, and span list.
  4. Continue in Explore → Endpoints and Explore → Errors.

Add an SLO and one policy​

  1. Open Services → SLO. Choose the service and environment. Set Availability (share of non-error requests), or Latency (P95 below threshold) / Latency (P99 below threshold) with a positive millisecond threshold. The target rate is greater than 0 and at most 100. Choose Rolling 7 days, Rolling 30 days, or Calendar month. An empty endpoint means the whole service.
  2. Open Events → Policies and create a policy. Enter the policy name and alert name. Choose a service option that already includes the environment, and choose the organization. The endpoint can be left empty.
  3. Choose one of error rate, P95, P99, throughput, or no-traffic. Set the evaluation interval, aggregation window, and aggregation method, then set comparators and thresholds for Critical, Error, and Warning. Assignees are optional. When set, they must belong to the policy organization and must not be disabled.
  4. Enable the policy and check Events → Alerts. The chart is the evidence from evaluation time, not a chart rebuilt from live traces.

Verify​

  • The service list and topology show the new call path.
  • Trace detail expands to spans, with sensitive fields masked.
  • The home page or service page shows whether the SLO is met.
  • An empty alert can be claimed or assigned. The current assignee can reassign. The empty alert or the current assignee can close it manually.
  • If notification is configured, confirm delivery in a channel already enabled in System Management. An Alert Center copy is not written back into APM.