Private Deployment
BabelBird private deployments run in an enterprise data center, private cloud, government/industry cloud or controlled cloud platform. They are designed for organizations with clear requirements for data location, network boundaries, identity, security policy, storage scale and system integration. A deployment can grow from POC or standard mode to high availability, K8S, dual data centers and multi-network Data Ferry.
This page helps select a route and links to detailed guides rather than duplicating every configuration chapter.
Typical Use Cases
- Files and databases must remain on enterprise-controlled infrastructure.
- Services operate in an intranet, secure zone or isolated network.
- AD/LDAP, WeCom, DingTalk, Feishu, OA or another SSO system must be integrated.
- Existing NFS, S3, object storage, backup or disaster-recovery platforms must be reused.
- High availability, dual data centers, Xinchuang, Data Ferry or dedicated AI is required.
- API integration, workflow customization, white label or long-term maintenance is required.
System Components
| Layer | Typical Components | Purpose |
|---|---|---|
| Access | Browser, Windows/macOS/Kylin clients and mobile apps | User access, sync, mapped drive and local editing |
| Entry | Nginx, reverse proxy, SSL and load balancer | Unified entry, certificates and traffic distribution |
| Application | BabelBird application servers and transcoding | File business logic, permissions, Web, preview and conversion |
| Data | MySQL-compatible database, optional Redis and search | Business data, cache and full-text index |
| File storage | Local disk, NFS, S3, MinIO or third-party object storage | File content, redundancy, capacity and expansion |
| Integration | SSO, OA, WeCom/DingTalk/Feishu, APIs and Data Ferry | Identity, business systems and cross-network exchange |
Component count depends on users, concurrency, file scale, availability targets and licensed modules. POC and standard deployments below 1000 users may colocate the database with the application. Standard deployments with 1000 or more users use one dedicated database server and one dedicated search server. Redis and independent monitoring are normally optional for high availability, K8S or dual-data-center designs.
Choose A Deployment Route
| Requirement | Recommended Route |
|---|---|
| Functional validation or short trial | POC Deployment |
| Small or medium production environment | Standard mode under Common Deployment Modes |
| Redundant application, database and storage | High-Availability Deployment |
| Continuous capacity growth or existing storage | Object Storage And Erasure Coding |
| Two data centers or multiple availability zones | Dual-site model in the high-availability guide |
| Secure and non-secure network exchange | Data Ferry Deployment |
| Domestic CPU, OS or database | Xinchuang Support |
| Deep business-system integration | Third-Party Services and Custom Development |
Implementation Path
- Determine licensed users, concurrency, file volume, annual growth, transcoding and AI load.
- Select standard, high availability, K8S, dual-site, cloud or POC deployment.
- Confirm application, database, search, transcoding and storage roles and quantities.
- Prepare domains, certificates, networks, ports, hosts and storage.
- Deploy, initialize, import organization and permissions, and integrate SSO and business systems.
- Validate transfer, permissions, search, preview, backup, failover and audit logs.
- Complete training and define upgrade, backup, monitoring and incident-response processes.
Deployment Guides
| Topic | Coverage |
|---|---|
| Configuration Advisor | Generate architecture and server suggestions by user count and mode |
| Server Requirements | Application, database, search, transcoding and storage sizing |
| System Architecture | Relationships between clients, services, databases, search and storage |
| Performance And Scalability | Cache, load balancing, queues, clusters and horizontal scaling |
| Ports And Network | External access, internal communication and deployment preparation |
| Common Deployment Modes | All-in-one, standard, multi-server, virtualized and cloud deployment |
| High Availability | Standby applications, databases, search, K8S, storage and dual sites |
| Object Storage | MinIO, S3, erasure coding, capacity and expansion |
| Implementation Process | Environment checks, installation, initialization, training and acceptance |
| Third-Party Services | Online Office, CAD, OCR, signatures, DLP and AI |
| Data Ferry | Cross-network transfer, tokens, direction, approval and audit |
| Custom Development | API integration, workflow, interface changes and OEM |
| Xinchuang Support | Domestic CPU, Linux, databases and clients |
Version And Delivery Boundary
Components, capabilities and ports vary by software version, licensed modules, third-party products and enterprise security policy. The approved deployment plan, server list, port matrix, license list and acceptance criteria are authoritative for each project.