Managing Runs¶
Check run status¶
tracker status RUN_NAME
Prints the current status of a run along with cluster, timing, and error details (if any).
my-run-abc123: running
cluster: pltf-dev
started: 2026-04-22T10:00:00Z
Exit codes:
| Code | Meaning |
|---|---|
0 |
Run found and status printed |
1 |
Authentication, not-found, or server error |
Stop a run¶
tracker stop RUN_NAME
Sends SIGTERM to the runner pod, which the workflow runner converts to an abort — task pods are cleaned up via Kubernetes ownerReferences.
Idempotent: if the runner pod has already terminated, exits 0 with a no-op message.
Authorization: You can stop a run if you are the run's owner, a cluster admin for the run's cluster, or are a global admin.
Exit codes:
| Code | Meaning |
|---|---|
0 |
Runner pod signaled, or already terminated (no-op) |
1 |
Authentication, permission, not-found, or server error |
View logs¶
tracker logs RUN_NAME
Prints the workflow runner log for a run. By default, tails the last 64 KB — enough to see recent activity without downloading the full log.
Use -full to download the entire log:
tracker logs RUN_NAME -full
Note
Logs are fetched directly from S3 using your ambient AWS credentials. Set AWS_PROFILE to the profile for the run's cluster (e.g. pltf-dev) before running. Works for in-progress runs as long as the runner log file has been written.
Wait for completion¶
tracker wait RUN_NAME [-timeout SECONDS] [-poll SECONDS]
Blocks until the run reaches a terminal status, printing RUN_NAME: STATUS (elapsed=Ns) to stderr on each poll.
| Option | Default | Description |
|---|---|---|
-timeout |
43200 (12 hours) | Maximum time to wait |
-poll |
30 | Seconds between status checks (minimum: 5) |
Exit codes:
| Code | Meaning |
|---|---|
0 |
Run succeeded |
1 |
Run failed, or authentication/not-found error |
124 |
Timed out before reaching a terminal status |