Skip to content

Fortified IDDocs

Recovery diagnostics

Recover interrupted AS4 transfers, handle database outages and plan access point upgrades or rollback safely.

Start recovery by identifying the affected transport ID and the last durable outcome. Preserve the database, release artifacts and instance configuration while working out what happened.

  1. Request status using the original transport ID and record the state, attempt count and receipt headers. Use the tenant-scoped journal to match the last attempt and outcome.
  2. Address the underlying cause using the operational reason code. After a process interruption, the worker can reclaim a persisted in-flight attempt. For eligible transient failures, allow the scheduled retry to run.
  3. If the result is uncertain, reconcile with the receiving side and its available evidence before sending a new package. When retrying the same submission through Interface C, reuse its original idempotency key. A new submission can create a new transport delivery. Integration

Check the database service, connectivity, TLS trust, runtime credentials and free storage. After connectivity returns, wait for private readiness and check existing dispatch and inbox state. Investigate pending work using the original IDs. Confirm that new synthetic traffic completes and that outstanding packages can still be retrieved.

  1. Keep the source and target release artifacts and the matching instance configuration. Stop the access point in an orderly way, create a database backup, and check that the backup restores.
  2. Run the selected release’s explicit schema migration with the migrator login. Start the runtime with its restricted login and check readiness, pending package retrieval and a new AS4 exchange.
  3. For rollback, choose the procedure for the exact source and target releases. If the previous runtime can use the current schema, preserve the schema history and transport data when switching the executable back. If the releases require a database restore, account for transactions after the backup before restoring.

The runtime does not apply automatic down migrations. Keep signed signals and journal rows with the corresponding dispatch and inbox state during recovery. Databases · Logging and audit