Backups and export

Two layers to back up:

1. The platform repository - artefacts, user workspaces, system files.
2. The data - rows in the data sources backing the application.

Exporting the repository

Use the platform's IRepositoryExporter to produce a ZIP of /registry/public/..., /users/.../workspace/..., and any other repository paths. Two ways:

• IDE: right-click a project / workspace in the Projects view -> Export -> Download ZIP.
• REST: GET /services/ide/transport/zip?path=/<path> returns a ZIP stream.

The ZIP is restorable through the IDE's import flow or via the matching POST /services/ide/transport/import endpoint.

Backing up the data

The platform itself does not snapshot the underlying database. Use the database's native backup tools:

• Postgres: pg_dump (logical) or filesystem-level snapshots.
• MSSQL: BACKUP DATABASE.
• HANA: HANA Backint / native backup.
• H2: copy the file under <repo>/dirigible/h2/ while the JVM is stopped.

For multi-tenant deployments back up SystemDB and the per-tenant data sources independently.

Disaster-recovery flow

1. Stop the runtime (or cordon ingress for zero-write).
2. Snapshot the database.
3. Export the repository ZIP.
4. Store both with the same timestamp - the repository state and the database state must be consistent.

A coupled restore replays both at the same timestamp.

CI-driven export

A small scheduled job can produce ZIP snapshots:

import { Repository } from "@aerokit/sdk/platform";

const bytes = Repository.exportPath("/registry/public");
// store bytes wherever - S3, blob storage, etc.

See also

• Projects view
• Data transfer