Skip to main content

Eloq Operator Introduction

Eloq Operator is a Kubernetes operator that automates the deployment and management of EloqKV and EloqDoc database clusters on Kubernetes. It follows the Kubernetes Operator pattern and uses Custom Resource Definitions (CRDs) to provide declarative cluster configuration.

Getting Started​

Prerequisites​

Before installing Eloq Operator, ensure you have:

  • Kubernetes cluster (v1.28 or higher) on AWS EKS, Baidu Cloud CCE.
  • kubectl (v1.28+) and helm (v3.0+)
  • Cloud provider CLI for cloud-specific features

Installation​

Step 1: Install Operator on Public Cloud Kubernetes

Depending on your cloud platform:

Step 2: Deploy Database Clusters

Storage Types​

Eloq Operator manages three types of storage for optimal performance and cost:

Block Storage (Persistent Volumes)​

Purpose: Raft consensus logs and metadata

  • Performance: High IOPS, low latency
  • Durability: Data persists across pod restarts
  • Providers: AWS EBS, Baidu CDS
  • Use Case: Transaction logs, critical metadata

Example:

spec:
  dataStore:
    pvc:
      spec:
        accessModes:
          - ReadWriteOnce
        resources:
          requests:
            storage: 500Mi
          limits:
            storage: 3Gi
        volumeMode: Filesystem
        # storageClassName: gp3  # AWS EBS gp3 type

Local Storage (Instance Store)​

Purpose: Hot data cache and high-performance operations

  • Performance: Ultra-low latency with NVMe SSDs
  • Durability: Ephemeral (data lost on pod restart)
  • Providers: AWS i3/i4i instances, Baidu bcc.l5d instances
  • Use Case: Active data cache, temporary computations

Example:

spec:
  dataStore:
    ephemeral:
      spec:
        accessModes:
          - ReadWriteOnce
        resources:
          requests:
            storage: 5Gi
          limits:
            storage: 10Gi

Object Storage (S3-Compatible)​

Purpose: Long-term data persistence and cold storage

  • Performance: Optimized for throughput
  • Durability: 99.999999999% (11 9's)
  • Providers: AWS S3, Baidu BOS, any S3-compatible storage
  • Use Case: SST files, transaction log archives, backups

Example:

spec:
  storage:
    objectStorage:
      type: s3
      bucket: eloq-data-bucket
      region: us-west-2

How Storage Tiers Work Together​

Write: Client → Local SSD (cache) → Object Storage (persistent)
                    ↓
              Block Storage (metadata)

Read (Hot):  Client → Local SSD → Return
Read (Cold): Client → Local SSD (miss) → Object Storage → Cache → Return

Benefits:

  • Hot data served from fast local SSDs
  • Cold data automatically tiered to cost-effective object storage
  • All data durably persisted in object storage
  • Metadata and logs stored in reliable block storage