Telemetry
VectorFlow can send anonymous, aggregate usage stats to a centralised receiver to help the maintainer understand how it's used and prioritise improvements. This is opt-in — VectorFlow does not send any data unless you explicitly turn it on.
What is collected
When telemetry is enabled, VectorFlow sends a single JSON payload once per day to https://pulse.terrifiedbug.com/api/v1/ping. The payload contains:
| Field | Source | Why |
|---|---|---|
schema_version | hardcoded 1 | lets the receiver evolve without breaking old clients |
instance_id | random ULID generated when you first opt in | a stable identifier for your installation. Anonymous — not linked to any user, organisation, or hostname. |
instance_created_at | timestamp of your first opt-in | cohort tracking — lets us see retention over time |
sent_at | current timestamp | when the heartbeat was sent |
vf_version | process.env.VF_VERSION (or "unknown") | which version of VectorFlow you're running |
agent_count | row count of the VectorNode table (fleet agents managed by VectorFlow) | scale signal — how many fleet agents are managed |
pipeline_count | derived from the Pipeline table — active = isDraft: false with deployedAt set, paused = isDraft: false with deployedAt null, draft = isDraft: true | { active, paused, draft } — scale and lifecycle signal |
auth_method | derived from OIDC issuer setting | "credentials" or "oidc" — informs which auth path matters |
deployment_mode | process.env.VF_DEPLOYMENT_MODE (or "unknown") | "docker", "helm", "bare", or "unknown" |
What is NOT collected
- Hostnames, IP addresses (the receiver's IP-based country lookup is server-side only — your IP is never stored, only a hash for rate limiting)
- Pipeline names, configurations, or VRL code
- User emails, names, or identifiers
- Source/sink endpoint URLs (Kafka brokers, S3 buckets, syslog servers, etc.)
- Any data flowing through your pipelines
Example payload
{
"schema_version": 1,
"instance_id": "01HX0000000000000000000000",
"instance_created_at": "2026-04-25T10:00:00.000Z",
"sent_at": "2026-04-26T03:42:00.000Z",
"vf_version": "1.4.2",
"agent_count": 5,
"pipeline_count": { "active": 12, "paused": 2, "draft": 3 },
"auth_method": "credentials",
"deployment_mode": "docker"
}How to opt out
In VectorFlow, go to Settings → Telemetry and turn the toggle off. Telemetry stops immediately. Your instance ID is preserved in case you re-enable later (so you appear as the same anonymous instance, not a new one).
You can also choose "No thanks" during initial setup if you want telemetry off from day one.
Receiver source
The telemetry receiver is private. Its data-handling code is documented in its own PRIVACY.md. No telemetry data is shared per-instance publicly — only aggregate counts on a community stats dashboard.