Quick answer
Capture changing files and database data at a state the application can recover together. Coordinate writes, jobs and external dependencies with their owners. A storage snapshot or an InnoDB transaction can help, but neither alone proves that every application component matches.
Define the recovery point the application needs
For a changing application, files and database data captured at different times may not form a usable recovery point. Identify the components and write paths, then agree an acceptable recovery point and the pause or synchronisation method.
Identify every component to recover together: application code and version, uploaded files, database, configuration references and any separate object storage or service state. Do not place secrets in a public backup location. A DirectAdmin account backup may be useful, but the existence of an archive alone does not prove that all changing components form one consistent application state.
Capture a bounded state with the operator
For a managed VPS, ask the operator to plan the database capture and any guest-level snapshot. The application owner can announce a maintenance or read-only window, stop new writes through the approved application procedure, and record the time at which jobs were paused. Avoid simply switching off the server; queued or external writes may still need reconciliation. The operator should state exactly which files and database objects are included and whether the capture method requires a lock or pause.
MariaDB documents mariadb-dump --single-transaction as providing a consistent state for InnoDB tables under its stated conditions. It does not guarantee a consistent dump for MyISAM or MEMORY tables, and concurrent schema-changing statements can invalidate the result. The operator must check the actual table engines and workload before choosing that option. Even a transactionally consistent database dump does not automatically align with files copied at another time.
Example: an upload and its database record
An upload written at 10:01 may be missing from a file copy captured at 10:00, even when a 10:02 database dump contains its record. Restoring that pair can show a broken attachment. Capture both under an agreed write pause or another application-supported consistency method, then verify the file and its record together.
If a job or external integration continues writing during the pause, record the exception and reconcile it. Do not describe the whole recovery set as consistent solely because the database dump completed successfully.
Record the capture boundary and artefacts
Record the application release, database name or identifier, relevant backup job reference, start and completion times, read-only window and the person who approved it. Confirm that each artefact is complete, readable by the authorised recovery team and stored under the agreed retention and access controls. Do not treat a successful backup job status as a restore test. If writes resumed before every required component was captured, document the gap rather than silently labelling the set consistent.
Rehearse restoration before relying on the backup
Where an authorised separate destination is available, have the operator restore the captured set without touching production. The application owner can then check a representative recent transaction, an uploaded file and a database-backed page. Compare counts or known record identifiers with the agreed capture point, while avoiding writes into the restored copy or live external services. If a component is missing, a schema version disagrees or the database cannot start, repeat the capture with a corrected method before a migration. Preserve the original artefacts for investigation.
Frequently asked questions
Is a VPS snapshot automatically application-consistent? No. Storage-level capture alone does not establish that the application, database and external writes reached a coordinated state.
Can I use a database dump made while orders continue? It depends on the database engines, dump options, schema activity and whether the application’s files or external state also change. Agree the method with the operator.
Sources and technical references
Related guides
Was this guide helpful?
If you accepted optional analytics, your Yes/No answer and this guide’s page address are sent to Google Analytics. No written feedback is collected. Read our privacy statement or change your choice in .
Need help with your service?
Open your client area or contact Novasean. Describe the symptom and checks you have performed. Keep passwords, recovery codes and private keys out of your message.