Service: Strom - EyevinnOSC/community GitHub Wiki

Strom

Strom (serviceId eyevinn-strom) is an open source GStreamer-based GUI and flow engine with a visual, web-based editor for building and managing streaming media pipelines. In Eyevinn Open Source Cloud (OSC) it is a long-running service instance with up to 10 GiB of storage per instance. License: Apache-2.0.

What it can do

  • Wire together sources, filters, encoders, muxers and outputs in a visual editor without writing GStreamer command lines.
  • Use WebRTC WHIP Output and WHEP Input/Output blocks for real-time streaming.
  • Prototype live encoding and distribution workflows.
  • Manage flows through a REST API.
  • Store flows in a local JSON file (default) or in PostgreSQL for persistent storage shared across instances.

Creating an instance

Strom has no required dependencies. PostgreSQL is optional: by default Strom stores flows in a local JSON file; for production use or to share flows across instances, connect a PostgreSQL database.

  1. Optional: create a PostgreSQL instance (birme-osc-postgresql, see PostgreSQL) with create-service-instance, giving a name and PostgresPassword, and wait until it runs. Read the IP and port with describe-service-instance. The URI is postgresql://postgres:<password>@<IP>:<PORT>/strom. Because it embeds the password, store it as a service secret with create-service-secret (serviceId: "eyevinn-strom", for example name dburl).
  2. Create the Strom instance with create-service-instance using serviceId: "eyevinn-strom". The instance name must be alphanumeric (underscores allowed) and at most 20 characters.

Minimal (JSON file storage):

{ "name": "mystrom" }

With PostgreSQL:

{ "name": "mystrom", "DatabaseUrl": "{{secrets.dburl}}" }
  1. Verify with wait-for-service-ready or get-instance-status, then read the URL with describe-service-instance.

OSC CLI example:

# Without a database (JSON file storage)
osc create eyevinn-strom mystrom

# With PostgreSQL backend
osc create eyevinn-strom mystrom \
  -o DatabaseUrl="postgresql://postgres:mypassword@<IP>:<PORT>/strom"

Options

Option Required Type Meaning
name Yes string Name of the instance. Letters, digits and underscore only; on OSC at most 20 characters.
DatabaseUrl No string PostgreSQL connection URL for storing flows and blocks, for example postgresql://user:pass@host:5432/strom. When set, Strom uses PostgreSQL instead of JSON file storage, runs schema migrations on startup, and several isolated instances can share one PostgreSQL server. Contains a password, so store as a secret and pass as {{secrets.NAME}}.
IceServers No string ICE server (STUN/TURN) configuration for WebRTC connections used by WHIP/WHEP blocks. Needed for WebRTC connectivity especially when clients are behind NAT or firewalls.

A {{secrets.NAME}} reference must be the whole field value. Overwriting a secret does not restart a running instance.

PostgreSQL options (birme-osc-postgresql)

Option Required Type Meaning
name Yes string Instance name, at most 20 characters.
PostgresPassword Yes string, sensitive Password for the PostgreSQL superuser. Example: {{secrets.postgrespwd}}.
PostgresUser No string Superuser name; postgres if not provided.
PostgresDb No string Name of the default database created at first start; the user name if not specified. Create or name a database strom to match the URI above.
PostgresInitDbArgs No string Extra arguments passed to initdb when the cluster is initialized.
PostgresInitDbSql No string SQL executed during database initialization.

Using the instance

Open the instance URL. The visual editor shows a canvas where you:

  1. Click Add block to add a source, filter, encoder, muxer or output block.
  2. Connect blocks by dragging from an output pad to an input pad.
  3. Click Start to launch the pipeline and monitor it in real time.

For a simple test pipeline, add a videotestsrc source and connect it to an autovideosink output, then click Start.

Strom exposes a REST API for managing pipelines:

# List all flows
curl https://<your-instance-url>/api/flows

# Start a flow by ID
curl -X POST https://<your-instance-url>/api/flows/<flow-id>/start

Limitations

  • Storage is limited to 10 GiB per instance.
  • With the default JSON file storage, flows are not shared across instances; use PostgreSQL for that.

Resources

⚠️ **GitHub.com Fallback** ⚠️