Blogacrynite.com

How Acrynite works

Acrynite separates backup configuration, backup execution and recovery into clear steps.

Acrynite separates backup configuration, backup execution and recovery operations. A typical lifecycle moves through six steps.

The workflow

1

Configure

Register the PostgreSQL database and the backup destination.

2

Back up

Run a logical or physical backup.

3

Record

Acrynite records the run in its catalog and stores a manifest beside the artifact.

4

Automate

Create a recurring backup schedule.

5

Archive WAL

Configure PostgreSQL to send WAL segments to the archive destination.

6

Recover

Select a backup or target time, inspect the plan, and execute recovery.

Conceptual architecture

The Acrynite CLI holds the configuration and catalog. It connects to a source PostgreSQL cluster to read data, and to S3 (or a local path) to store backup artifacts and WAL. For recovery, it retrieves those artifacts and replays WAL into an isolated Docker PostgreSQL container.

+-----------------------+ | Acrynite CLI | | configuration/catalog | +-----------+-----------+ | +---------------+---------------+ | | +------v-------+ +------v-------+ | PostgreSQL | logical dump | S3 storage | | source | -------------> | backups/WAL | | cluster | physical base | | +------+-------+ +------+-------+ | | WAL archiving retrieve on | recovery +---------------+---------------+ | +---------v----------+ | isolated Docker | | PostgreSQL (restore)| +--------------------+

The diagram is conceptual

Exact data movement and requirements differ between logical and physical workflows. A logical dump exports database objects through pg_dump; a physical base backup copies cluster files through pg_basebackup and pairs with WAL for point-in-time recovery.