Conflux Stack

Platform
Pricing
docs

Layer 09 · Platform Operations

Govern tenants, releases, and lifecycle from one operations layer

Govern multi-tenant deployments with feature flags, observability, data portability, backup, and versioning. Operations modules manage the lifecycle of everything above them.

Feature FlagsTenant ManagementObservabilityImport & ExportBackup & RestoreVersioning

Why this layer

Built into the platform, not bolted on

Every module in this layer solves the same class of problem the same way — regardless of who implements it.

01

Multi-tenant by default

Tenant provisioning, isolation, and per-tenant configuration are platform modules — not custom middleware every application reimplements.

02

Ship safely with flags and versions

Feature flags, schema versioning, and API version pinning deploy through CLI. Gradual rollouts and rollback paths are first-class.

03

Operate with unified observability

Audit log, analytics (DuckDB over SQLite), and Console Monitoring (health + job-queue pressure). User-defined alerting rules are not shipped yet.

Canonical implementation

One path for every author

Junior developers, senior engineers, and AI agents produce the same standard output.

How Conflux does it

Operations modules sit below every other layer

Feature flags evaluate through SDK with local cache. Backups run on schedule with retention policies. Import/export validates against schema before commit.

Canonical path

conflux.flags.evaluate() scoped to tenant, user, or environment

Backups and restore through CLI with dry-run preview

Observability spans auth, data, runtime, and communications

Schema and API versions pinned per deployment environment

What you get

Provision and isolate tenants without custom middleware

Roll out features gradually with SDK-cached flag evaluation

Recover tenant data with point-in-time backup and restore

Debug production issues from one cross-layer observability view

01 · Platform Operations

Roll out features with confidence

Gradual rollout and experimentation scoped to tenant, user, or environment. Flag evaluation through SDK with local cache for performance.

Feature Flags · One canonical path

SDK

conflux.flags.isEnabled('new-checkout')

CLI

conflux flags set new-checkout 10%

Config

flags:
  - new-checkout

What's possible with Feature Flags

01

Gradual rollout and experimentation

02

Flags scoped to tenant, user, or environment

03

Evaluation through SDK with local cache

02 · Platform Operations

Multi-tenant isolation from day one

Tenant provisioning, configuration, and administration through CLI and console. Per-tenant settings and billing hooks without custom middleware.

Tenant Management · One canonical path

SDK

await conflux.tenants.current()

CLI

conflux tenants create acme

Config

tenants:
  isolation: strict

What's possible with Tenant Management

01

Multi-tenant isolation and administration

02

Tenant provisioning through CLI and console

03

Per-tenant configuration and billing hooks

03 · Platform Operations

Logs, metrics, and traces in one place

Unified audit, DuckDB analytics, and Console Monitoring across auth, data, runtime, and communications. Custom alerting rules are planned — not shipped.

Observability · One canonical path

SDK

conflux.observability.trace('checkout')

CLI

conflux logs tail --service api

Config

observability:
  alerts: [error-rate]

What's possible with Observability

01

Logs, metrics, and traces across all layers

02

Unified dashboard for platform and app health

03

Audit + analytics + monitoring health surfaces

04 · Platform Operations

Move data between environments safely

Bulk import with schema validation and export formats for backup and compliance. Data portability without one-off migration scripts.

Import & Export · One canonical path

SDK

await conflux.data.export({ format: 'json' })

CLI

conflux data import orders.csv

Config

import:
  validate: schema

What's possible with Import & Export

01

Data portability and migration between environments

02

Bulk import with validation against schema

03

Export formats for backup and compliance

05 · Platform Operations

Point-in-time recovery for tenant data

Host backups via scripts/backup-conflux.sh; Conflux Admin (apps/admin) + CLI show status. Restore is a VPS runbook operation (verify-backup-conflux.sh), not an in-Console action.

Backup & Restore · One canonical path

SDK

await createAdminClient({ url, adminToken }).ops.backups()

CLI

conflux backup status

Config

# CONFLUX_DATA_DIR + nightly backup-conflux.sh

What's possible with Backup & Restore

01

Point-in-time recovery for tenant data

02

Scheduled backups with retention policies

03

Restore through CLI with dry-run preview

06 · Platform Operations

Schema and API versions that ship together

Backward-compatible migration paths with version pinning per deployment environment. Breaking changes are explicit, documented, and deployable.

Versioning · One canonical path

SDK

import { api } from '@conflux-stack/sdk/v2'

CLI

conflux version pin v2

Config

version:
  api: v2

What's possible with Versioning

01

Schema and API version management

02

Backward-compatible migration paths

03

Version pinning per deployment environment

Explore

Continue through the platform

Nine layers, one cohesive stack. Browse adjacent layers or return to the full registry.

01 · Identity

02 · Data

03 · Runtime

04 · Communication

05 · Intelligence

06 · Automation

07 · Integration

08 · Experience

09 · Platform Operations

Previous layer

Experience Layer

View full platform registryTalk to the team