| name | exporting-traffic-visualization-flows |
| description | Export large batches of summarized traffic-visualization flow rows from CloudTower as CSV. Best suited when you need broad cluster-wide flow snapshots rather than deep interactive TQL. Use when the user needs flow exports for downstream analysis or a plain file download.
|
Exporting traffic visualization flows
Common cases: inspect or summarize the CSV with your own tools, or deliver the file as-is. Prefer export-traffic-visualization-flows.py --help for flags and defaults.
After you download the CSV on the async path, you must delete the server-side async request (see Workflow).
Quick start
Export one hour of traffic visualization flow data for a specific cluster associated with an observability service.
Always pass these three flags: --obs (observability instance), --from_time, and --to_time. They are required to scope the export; do not omit them when running the script for this use case.
python3 export-traffic-visualization-flows.py \
--env=/path/to/.env \
--obs=my-observability-service-name \
--clusters=my-cluster-name \
--from_time=2026-01-01T11:00:00Z \
--to_time=2026-01-01T12:00:00Z
Use real values. Timestamps are RFC3339 (timezone Z or numeric offset). --obs accepts a single name, id, or management IP (no comma-separated list).
Workflow
- Authenticate against CloudTower (e.g.
.env with TOWER_URL and token or username/password).
- Scope the run: observability instance, clusters, UTC window, optional filters (
--help).
- If the export is asynchronous: poll until ready, download the file, then immediately delete that async request using the script’s delete operation (
--help for the flag). Treat download + delete as one unit of work—do not stop after download, even when the next step is CSV analysis. Optional --sync-export with --output performs download and delete in one process; reserve it for very small result sets. Tune polling and timeout via --help if needed.
Examples: status, download, delete
After you start an async export (without --sync-export), the JSON response includes:
id — async request id (UUID). Use this for --request-status and --delete-request.task_status.id — export task id (long hex string). Use this for --download-task once task_status.ready is true.
Replace placeholders with values from that response. Each command needs --env and --obs (same instance as the export).
Poll status (repeat until task_status.ready is true and there is no error):
python3 export-traffic-visualization-flows.py \
--env=/path/to/.env \
--obs=my-observability-service-name \
--request-status=REQUEST_UUID
Download CSV (only after the task is ready):
python3 export-traffic-visualization-flows.py \
--env=/path/to/.env \
--obs=my-observability-service-name \
--download-task=TASK_ID \
--output=/path/to/flows.csv
Delete the async request (required after a successful download; use the request UUID, not the task id):
python3 export-traffic-visualization-flows.py \
--env=/path/to/.env \
--obs=my-observability-service-name \
--delete-request=REQUEST_UUID
Progress checklist
- [ ] Credentials and Tower base URL ready
- [ ] Instance, clusters, time window (and filters if needed)
- [ ] Async path: status, download, **then delete request** (never skip delete after download; or use `--sync-export` only when data is tiny)
