Frequently asked questions

Migration, licensing, support, sovereignty and operations questions, answered.

Do we have to replatform to use Streamtime?
No. Support plans cover existing Kafka estates as-is — managed, self-hosted, or on any distribution. Every engagement also includes free migration services and $10,000 of consulting from Kafka experts, so you can move at your own pace.
What infrastructure does Streamtime run on?
Bootstrap directly on AWS, Azure, GCP, DigitalOcean, Oracle Cloud or Civo. Or bring your own Kubernetes: any OpenShift or CNCF-conformant cluster (1.32+) with a load balancer and a default CSI storage class. Data Center and edge deployments are supported.
Which workloads are supported?
Multi-tenant streaming (Apache Kafka and Confluent Platform with virtual clusters, Kafka Connect with hundreds of connectors including built-in Debezium, MirrorMaker for autonomous migrations and DR), stream processing (Apache Flink, Kafka Streams) and real-time analytics (ClickHouse, Apache Iceberg compatible) — on one fleet with one identity model and one operations model.
How is Streamtime licensed?
By usage, not reservation. Streaming workloads are metered in Kafka Units: 1 KU = 5 MB/s write + 15 MB/s read. Apache Flink and Kafka Streams are priced from state store size and parallel processing slots actually occupied. Bring-your-own-license models apply your existing OSS or commercial agreements directly.
What support plans are available for existing Kafka?
Silver ($12,000 per cluster, business-hours support), Gold ($24,000 per cluster, 24×7 across up to 3 environments) and Platinum (mission-critical production, engagement-specific terms). See the pricing page for details.
Can Streamtime run in air-gapped environments?
Yes. Connected, restricted-egress and true air-gap deployments share the same APIs, operators and workflows. In a true air-gap, the entire management plane moves inside your boundary with local registries, offline bundles and self-healing control planes.
How does support work without an air-gapped environment?
Air-gapped deployments get an insurance-style cover model: local services designed to self-heal against known failure modes, offline bundles carrying tested fixes, and support engagements through auditable, policy-controlled artifact exchanges rather than live tunnels into your enclave.
Is the TCO assessment really free?
Yes. We assess the total cost of your existing Kafka estate — anywhere it runs — and commit to a guaranteed price reduction in writing. If our proposal cannot beat your current spend, the assessment costs you nothing. Contact us to start.
Can AI agents operate the platform?
Yes. A Streamtime MCP server exposes the whole fleet — desired vs. actual state, metrics, recent changes, in-flight operations — to any MCP-capable agent, and the Streamtime AI assistant provides a purpose-built operations harness with guarded, verifiable actions.
How is capacity sized?
Empirically, not by rules of thumb. Clusters are benchmarked on your hardware — sustained throughput, latency envelopes and failure-mode headroom — and validated against your workload's declared intent before production use.
Where can I learn more?
Read the Sovereign Streaming whitepaper, browse docs.streamtime.ai, or contact us for a trial.