Skip to main content

Dynatrace Webhook

The Dynatrace webhook integration forwards Dynatrace Problems (incidents) directly into NudgeBee. When a problem fires or resolves in Dynatrace, NudgeBee automatically receives it, enriches it with correlated logs and traces, and creates a trackable incident.

Prerequisite: For full telemetry enrichment (auto-attached logs and traces), configure the Dynatrace Observability integration first.

How It Works​

Dynatrace Davis detects a problem
        │
        ▼
Dynatrace Workflow triggers (Davis problem)
        │
        ▼
HTTP Request action sends POST to NudgeBee
        │
        ▼
NudgeBee  /webhook/dynatrace
        │
        ├── Fetch full problem details from Dynatrace API v2
        ├── Match impacted entity to a Kubernetes workload
        ├── Auto-collect correlated logs (2 hrs before → 6 hrs after)
        ├── Auto-collect correlated traces (same window)
        └── Create NudgeBee incident with evidence attached

Step 1: Copy the Webhook URL from NudgeBee​

  1. In NudgeBee, navigate to Integrations > Webhooks
  2. Select Dynatrace
  3. Copy the webhook endpoint URL — it will look like: 
    https://<your-nudgebee-url>/webhook/dynatrace

Step 2: Configure a Workflow in Dynatrace​

  1. In Dynatrace, open the Workflows app (search for "Workflows" in the navigation bar).
  2. Click Add Workflow.
  3. Give the workflow a descriptive name (e.g., NudgeBee Problem Notification).

Configure the Trigger​

  1. Set the trigger type to Davis problem.
  2. Configure the trigger settings:
    • Problem state: select Active or Closed — this ensures the workflow fires both when a problem opens and when it resolves.
    • Event category: select All to capture all problem types, or pick specific categories (Availability, Error, Slowdown, Resource, Custom).
    • (Optional) Add entity tag filters to limit which problems are forwarded (e.g., only problems tagged env:production).
    • (Optional) Add a DQL matcher expression for fine-grained filtering (e.g., maintenance.is_under_maintenance == false to skip maintenance windows).

Add the HTTP Request Task​

  1. In the workflow canvas, click + (below the trigger node) to add a task.
  2. Select the HTTP Request action.
  3. Configure the task:
    • Method: POST
    • URL: paste the NudgeBee webhook URL from Step 1
    • Headers: add Content-Type with value application/json
    • Payload: paste the following:
{{ event() | to_json }}

This sends the entire Davis problem record to NudgeBee, including all entity metadata, Kubernetes context, severity, timestamps, and tags.

  1. Click Save to save the workflow as a draft, then click Deploy to activate it.

Reference: Dynatrace Workflows documentation

Custom Payload Template

If you prefer to send only specific fields instead of the full event, use this template in the payload:

{
  "event.id": "{{ event()['event.id'] }}",
  "display_id": "{{ event()['display_id'] }}",
  "event.status": "{{ event()['event.status'] }}",
  "event.name": "{{ event()['event.name'] }}",
  "event.category": "{{ event()['event.category'] }}",
  "event.start": "{{ event()['event.start'] }}",
  "event.end": "{{ event()['event.end'] }}",
  "event.kind": "{{ event()['event.kind'] }}",
  "k8s.workload.name": {{ event()['k8s.workload.name'] | to_json }},
  "k8s.namespace.name": {{ event()['k8s.namespace.name'] | to_json }},
  "affected_entity_names": {{ event()['affected_entity_names'] | to_json }},
  "entity_tags": {{ event()['entity_tags'] | to_json }},
  "dt.davis.impact_level": {{ event()['dt.davis.impact_level'] | to_json }}
}

Dynatrace Workflows use Jinja template expressions with the event() function to access problem fields. The | to_json filter ensures array values are serialized as valid JSON arrays.

Legacy: Problem Notifications (Dynatrace Classic)

If your Dynatrace environment does not have the Workflows app (e.g., older SaaS or Managed versions), you can use the legacy Problem Notifications approach:

  1. Go to Settings > Integrations > Problem notifications
  2. Click Add notification > Custom integration
  3. Set Display name to NudgeBee, paste the webhook URL, and enable Accept SSL certificate
  4. Under Payload, paste:
{
  "event.id": "{ProblemID}",
  "event.status": "{State}",
  "event.name": "{ProblemTitle}",
  "event.severity": "{SeverityLevel}",
  "event.start_time": "{ProblemDetailsJSONv2.startTime}",
  "event.end_time": "{ProblemDetailsJSONv2.endTime}",
  "k8s.workload.name": "{ProblemDetailsJSONv2.impactedEntities[0].name}",
  "dt.problem.url": "{ProblemURL}",
  "dt.tags": "{Tags}"
}
  1. Select an Alerting profile to choose which problems to forward, then click Save

Note: Legacy Problem Notifications use {Placeholder} syntax. Status values are OPEN / RESOLVED / MERGED (instead of ACTIVE and CLOSED in Workflows).

Step 3: Test the Workflow​

  1. In the Dynatrace Workflows app, open the workflow you created.
  2. Click Run to execute the workflow manually, or wait for a real Davis problem to trigger it.
  3. Check the Execution log in the workflow to verify the HTTP Request task completed with a 2xx status code.
  4. In NudgeBee, navigate to Incidents — a test incident should appear within a few seconds.
  5. Open the incident and confirm:
    • Problem title and severity are populated
    • Impacted workload is linked (if entity mapping succeeded)
    • Log and trace evidence is attached (if the Observability integration is configured)

Event Mapping​

Problem State → NudgeBee Status​

Dynatrace Problem StateNudgeBee Incident Status
ACTIVEfiring
CLOSEDresolved

Severity → Priority​

Dynatrace SeverityNudgeBee Priority
AVAILABILITYP1 — Critical
ERRORP1 — Critical
PERFORMANCEP2 — High
RESOURCE_CONTENTIONP3 — Medium
CUSTOM_ALERTP3 — Medium

Automatic Evidence Collection​

When NudgeBee receives a problem, it automatically enriches the incident:

  1. Problem details — Full metadata is fetched from the Dynatrace Problems API (/api/v2/problems/{problemId}): impacted entities, tags, start/end times, linked events.
  2. Workload matching — The impacted entity name is matched to a Kubernetes workload using progressive lookups: exact name match → label app.kubernetes.io/name → label app → substring match.
  3. Log evidence — Correlated logs are queried from Dynatrace Grail for 2 hours before to 6 hours after the problem start time, filtered to the matched workload.
  4. Trace evidence — Distributed traces for the same workload and time window are attached.
  5. Deduplication — Problems are deduplicated using the Dynatrace Problem ID as a unique fingerprint. Reopened problems update the existing incident rather than creating a duplicate.

Evidence collection (steps 3–4) requires the Dynatrace Observability integration to be configured for the same environment.

Prerequisites Checklist​

  • The NudgeBee webhook URL is publicly reachable from Dynatrace's Workflow execution environment
  • Dynatrace Observability integration configured (for log and trace enrichment)
  • Access token includes storage:logs:read and storage:spans:read scopes
  • (Optional) environment-api:problems:read scope added for richer problem metadata

Verify the Integration​

  1. Confirm the workflow test (Step 3) created an incident in NudgeBee.
  2. Wait for a real Dynatrace problem to fire, or trigger one manually (e.g., via a synthetic monitor or by running the workflow manually).
  3. Open the incident in NudgeBee and verify:
    • Status reflects the Dynatrace problem state (firing / resolved)
    • Priority maps correctly to the problem severity
    • Logs and Traces tabs show correlated evidence

Troubleshooting​

SymptomLikely CauseFix
No incident created after running workflowWebhook URL unreachable or HTTP Request task misconfiguredCheck the workflow execution log for HTTP errors. Verify the NudgeBee URL is publicly accessible.
Workflow does not trigger on new problemsTrigger misconfiguredVerify the trigger type is Davis problem with problem state set to Active or Closed. Check event category filters.
Workflow fires only on open, not resolveProblem state filter incompleteSet problem state to Active or Closed (not just Active) in the trigger configuration.
Incident created but no workload linkedEntity name doesn't match any K8s workloadCheck if the impacted entity name in Dynatrace matches a workload name or label in NudgeBee.
No logs or traces attachedObservability integration not configuredSet up the Dynatrace Observability integration first.
403 on problem enrichmentToken missing environment-api:problems:readAdd the scope to your token, or the integration falls back to DQL-based enrichment.
Duplicate incidents for the same problemevent.id missing in payloadEnsure your payload includes the event.id field (included by default with `event()
Payload fields render as literal textJinja syntax error in templateVerify expressions use double braces and the event() function. Check the workflow execution log for template rendering errors.