Executions
Track plan execution history, live rollout status, and per-job results
Executions
The Executions page shows the history and live status of every change plan execution across your fleet. Filter by status, search by ID or plan name, and drill into individual executions for per-job details.
Navigation
- Parent: Automation Platform
- Previous: Change Plans
- Next: Insights Dashboard
- Related: Insights Nodes
Executions List
Navigate to Automation Platform → Executions to see all executions. The table shows:
| Column | Description |
|---|---|
| Execution | Short execution ID (click to view details) |
| Plan | Linked change plan name and commit SHA |
| Status | Current execution status |
| Jobs | Succeeded / failed / partial / total job counts |
| Queued | When the execution was queued |
| Completed | When the execution finished (or "—" if still running) |
Status Filters
Filter executions by clicking a status button:
| Status | Meaning |
|---|---|
| Pending | Queued, waiting to start |
| Running | Actively rolling out to nodes |
| Succeeded | All jobs completed successfully |
| Failed | One or more jobs encountered errors |
| Partial | Some jobs succeeded, some failed |
| Canceled | Execution was stopped before completion |
Use the search bar to find executions by ID or plan name.
Execution Details
Click an execution ID to open the detail view. You see:
- Execution summary — status, rollout strategy, timing, who triggered the run, and the linked change plan
- Per-job breakdown — each targeted node's job status (queued, running, succeeded, failed, partial, canceled)
- Live event stream — real-time worker events while the execution is in progress; the view refreshes automatically on retry so stale logs from a prior attempt are cleared
- Job logs — detailed output for each job, including applied config changes and any errors
- Timeline — queued, started, and completed timestamps
Per-Job Results
Expand any job to inspect what the worker actually did on that host:
| Section | What It Shows |
|---|---|
| Config files written | Files created, updated, or deleted on disk. Expand a file to see per-key INI mutations — stanza, option, change type, and desired value. |
| Splunk actions | Post-config actions such as bundle apply, restart, or rolling restart. Failed actions expand automatically and show exit code, message, and stdout/stderr snippets. |
| Failure reason | When a job fails, a concise failure summary appears on the job row. |
When a job is still queued or running, the results section shows Waiting for the worker node to execute this job. When a job succeeds but every file was already up to date, you see All config files already up-to-date — no changes written to disk.
Retry and Cancel
On the execution detail page you can:
- Cancel a running execution to stop rollout
- Retry failed or canceled jobs — succeeded hosts are left alone; only jobs that did not complete are re-queued
You can also retry from the Plans list when a plan's latest execution ended in Failed, Partial, or Cancelled status.
Job Counts
The jobs column uses color-coded counts:
- Green — succeeded jobs
- Red — failed jobs (suffixed with
f) - Orange — partial jobs (suffixed with
p) - Gray — total jobs
For example, 3/1f/4 means 3 succeeded, 1 failed, out of 4 total.
