Sinks¶

A sink is a datasource that varTrack writes config values into. One push can fan out to multiple sinks simultaneously — just add multiple rules pointing at different datasources.

Available sinks¶

Sink	Use case
MongoDB	Feature flags, app config, tenant settings stored as documents
Redis	Low-latency config reads, real-time feature flags
ZooKeeper	Distributed coordination config, service discovery values
S3	Config files as objects, audit storage, batch pipelines
Kubernetes ConfigMap	App config for Kubernetes workloads
Helm	Helm values files, chart config driven by Git
Vercel	Frontend environment variables synced from Git
Linux Server	Write config files directly to remote servers via SSH

Multi-sink fan-out¶

Push once, write to all of them:

rules: [
  { platform: "github", datasource: "mongo",      file_name: "config.yaml", destination_template: "{env}-config" },
  { platform: "github", datasource: "redis",      file_name: "config.yaml", destination_template: "{env}:cfg" },
  { platform: "github", datasource: "zookeeper",  file_name: "config.yaml", destination_template: "/{tenant}/{env}" },
  { platform: "github", datasource: "s3",         file_name: "config.yaml", destination_template: "{tenant}/{env}/" },
]

All four writes run in parallel on each push.

Multiple instances of the same type¶

Use tag to run two MongoDB instances side by side:

datasources: [
  { mongo: { tag: "primary",  endpoint: "mongodb://mongo-primary:27017",  database: "app" } },
  { mongo: { tag: "replica",  endpoint: "mongodb://mongo-replica:27017",  database: "app" } },
]

rules: [
  { platform: "github", datasource: "mongo-primary", ... },
  { platform: "github", datasource: "mongo-replica",  ... },
]

The resolved datasource name is mongo-{tag} when a tag is set, or mongo when the tag is empty.

Destination template¶

Every sink supports destination_template to control where data lands using {env} and {tenant} placeholders. See the Destination Template page for the full reference.