Enterprise Support
A named engineer, prioritized handling and response windows you can plan around.
A person, not a queue
You're assigned a named support engineer who learns your topology, your connectors and your traffic shape — so you're never re-explaining your stack to a first-line agent at 3am. Escalation reaches the core maintainers directly.
Prioritized issue handling
Your reports jump the public queue. A confirmed defect on your critical path is triaged first, patched, and — where you run a pinned build — backported to your LTS line so you don't have to chase the tip of dev to get a fix.
Response you can plan around
Every severity carries a committed response target and a coverage window, set in your agreement — 24×7 for anything that stops production. Predictable enough to put in your own downstream SLAs.
Severity decides speed — you decide severity.
You classify the incident when you open it. We respond to the clock that severity sets.
| Severity | What it means | First response | Coverage |
|---|---|---|---|
| P1 · Critical | Production is down or unusable. Cluster won't form quorum, data plane stalled, total outage. | ≤ 1 hour | 24 × 7 |
| P2 · Major | Severely degraded, no workaround. Failover not completing, a sink stuck, throughput collapse. | ≤ 4 business hours | business hours |
| P3 · Minor | Degraded with a workaround, or a real bug. Non-blocking defect, elevated latency you can absorb. | ≤ 1 business day | business hours |
| P4 · Guidance | Question, config review, how-to. Architecture advice, upgrade planning, best-practice. | ≤ 2 business days | business hours |
The times above are a baseline — your support agreement commits the exact targets, coverage window and time zone. Higher tiers tighten P1 and extend 24×7 to lower severities.
What a support engineer will actually pick up.
Core runtime, HA quorum, Raft replication, the cluster bridge — the parts that carry your data.
Kafka, JDBC, MQTT, S3 and the multi-region event mesh, including the ones you bridge in.
Version-to-version upgrade planning, migration dry-runs, rollback rehearsal before you touch prod.
Live outage triage, root-cause on your logs and traces, and the fix — not just a workaround.
Topology, partitioning, retention and capacity sizing reviewed against your real throughput.
CVE notifications ahead of public disclosure, backported patches, config hardening guidance.
Stay on a version — and stay patched
- Pin a release. Run a designated LTS line without being forced to the tip.
- Security backports. Critical fixes land on your pinned version, not only on dev.
- Planned upgrades. You move when you choose, with a rehearsed migration path.
- End-of-life notice. Clear, dated windows — never a surprise cutoff.
Direct lines, not a portal maze
- Private channel. A shared Slack or Teams space with your engineer and the maintainers.
- Named escalation manager. One owner for P1s, accountable end-to-end.
- Onboarding review. We learn your deployment before the first incident, not during it.
- Periodic health checks. Regular reviews of capacity, upgrades and risk on your cluster.
Put a name and a clock behind your deployment.
Enterprise Support is part of a StreamFlow Enterprise agreement. Tell us your topology and the coverage you need, and we'll shape the SLA around it.