Skip to content
BoilStream

BoilStream documentation

directory
TL;DR

Connect to a deployment, send your first rows and query them with DuckDB SQL. These guides separate everyday data tasks from deployment administration and exact interface references. Start with a small, verifiable workflow, then choose the producer and consumption interfaces whose visibility, durability and retry rules fit your application.

ArchitectureClients connect to BoilStream; BoilStream manages the data
Clients connect to BoilStream; BoilStream manages the dataProducers send rows to BoilStream ingestion interfaces. SQL clients query through the server. Inside BoilStream, incoming rows feed recent storage and processing. Durable data is published to object storage. Live applications receive a separate SSE stream, not a durable change log.BoilStream serverProducersServices and collectorsSQL clientspsql · DuckDB · Flight SQLIngestionHTTP Arrow · Kafka ·FlightRecent dataStorage and processingSQL interfacesPermissions and queryexecutionObject storagePublished data filesLive applicationsSSE consumersRowsAccepted inputRead visible dataSQL / resultsPublish filesRead historyLive eventsClients connect to BoilStream; BoilStream manages the dataProducers send rows to BoilStream ingestion interfaces. SQL clients query through the server. Inside BoilStream, incoming rows feed recent storage and processing. Durable data is published to object storage. Live applications receive a separate SSE stream, not a durable change log.BoilStream serverProducersServices andcollectorsSQL clientspsql · DuckDB ·Flight SQLIngestionHTTP Arrow · Kafka ·FlightRecent dataStorage andprocessingSQL interfacesPermissions and queryexecutionObject storagePublished data filesLive applicationsSSE consumersRowsAccepted inputRead visible dataSQL / resultsPublish filesRead historyLive events
These are separate paths: accepting input does not confirm object publication, a query result or live-client processing.

Use an existing deployment

Section titled “Use an existing deployment”

Run your first workflow to connect, create a table, insert two events and read them through the same client.

Operate a deployment

Section titled “Operate a deployment”

Start with deployment configuration or Kubernetes. Review completion boundaries before choosing retention and recovery settings.

Look up an exact contract

Section titled “Look up an exact contract”

Use the reference directory for configuration fields, protocol limits, commands and metrics. Find complete datasets in Examples or diagnose a symptom in Troubleshooting.