Recover from a failed update

หมายเหตุ: บทความนี้ยังเป็นภาษาอังกฤษ เนื่องจากคำแปลภาษาไทยกำลังอยู่ระหว่างจัดทำ

Purpose

Return the framework to a known-good runtime while preserving evidence and database backups.

Framework administrators and deployment operators.

Access to application logs, the previous framework version, and the latest verified backup.

Record the update job error and backup path from System Maintenance.
Check available disk space and application/updater logs.
Windows: inspect the timestamped framework backup and run Update.cmd after correcting the cause.
Docker: verify whether the sidecar restored the previous container, then follow the manual rollback procedure.
Start the previous framework version and verify data before retrying.
Restore the .emubackup only if data verification fails or release notes explicitly require database rollback.
Restore .emu-secret.key or the configured EMU_SECRET_KEY_PATH separately. After returning to a version that supports SMTP, verify the connection and send a test message.

Do not mix framework files from multiple releases. Preserve the failed state until logs and backups have been collected.