Skip to main content

Store

A store is the central unit of execution in Clustron DKV.

All data and coordination operations happen inside a store.

What is a Store?​

A store represents:

  • A logical cluster
  • A container for data and coordination primitives
  • A boundary for consistency and lifecycle management

You always interact with Clustron through a store.

One Store, Many Nodes​

In distributed mode, a store is backed by multiple nodes:

Store (TestStore)
  ├── Node 1
  ├── Node 2
  └── Node N
  • Each node participates in the cluster
  • Data and coordination are shared across nodes
  • The system behaves as a single logical unit

Store in In-Process Mode​

In InProc mode, a store runs entirely inside your application:

Application
  └── Store (InProc)
  • No network communication
  • No external dependencies
  • Same API as distributed mode

Store in Distributed Mode​

In distributed mode, a store runs across machines:

Machine A → Node 1
Machine B → Node 2
Machine C → Node 3
  • Nodes communicate with each other
  • Clients can connect to any node
  • The cluster maintains shared state

Naming a Store​

Each store has a unique name:

TestStore
orders
cache

The name is used when:

  • Creating the store
  • Connecting clients
  • Resolving from configuration

Store as a Boundary​

A store defines important boundaries:

Data Boundary​

  • Keys exist within a store
  • Different stores do not share data

Coordination Boundary​

  • Locks, counters, and watches are scoped to a store
  • No cross-store coordination

Failure Boundary​

  • Failures and recovery happen within a store
  • Each store operates independently

Multiple Stores​

You can run multiple stores in the same environment:

orders-store
inventory-store
session-store

Each store is isolated and can be managed independently.

Why Stores Matter​

The store is what makes Clustron scalable and modular:

  • You can isolate workloads
  • You can scale specific stores independently
  • You can apply different configurations per store

What’s Next​

👉 Continue to InProc vs Remote to understand execution modes