Skip to content
Müşavir

Backup and Recovery

How much data could you lose if something goes wrong

The real answer to the backup question is two numbers: in the worst case how many hours of data can be lost, and how quickly the service comes back. This page describes the setup we run today and our targets for those two numbers, clearly marked as targets.

Write to us

The sequence followed during an incident

  1. 1

    Detection

    The problem is spotted through monitoring alerts or a user report, and its scope is determined.

  2. 2

    Notice

    Affected tenants are informed through in-app notifications and email.

  3. 3

    Restore

    The last healthy backup before the problem is selected, verified in a separate environment first, then put live.

  4. 4

    Closing report

    What happened, which time window of data was affected and the measures taken are shared with tenants.

))} )}

Backup setup and targets

RPO and RTO are targets, not contractual guarantees. Contract-specific targets can be discussed on the Enterprise plan.

TopicCurrent practice or target
Database backupTarget: a full backup every 6 hours
Backup locationBackups are kept in a storage service separate from the database service
RetentionTarget: backups from the last 24 hours available at any time and daily backups kept for 30 days
RPO (maximum accepted data loss)Target: 6 hours
RTO (time to restore service)Target: 8 hours
Restore testingTarget: restore a backup to a separate environment and verify it at least once a month
Encrypted second copy outside the clusterRoadmap

What we back up

The platform database holds all customer, invoice, journal, reconciliation, HR and task records, and our target is to back it up every 6 hours. Backups are written to a storage service separate from the database. However, since that service is in the same cluster, it is not on its own sufficient against a cluster-wide storage failure. Keeping an encrypted second copy of backups outside the cluster is therefore on our roadmap; we will update this page when it is done.

Targets and measurement

  • RPO target of 6 hours: Follows from the backup interval.
  • RTO target of 8 hours: The planned time to select a backup, verify it in a separate environment and put it live.
  • Restore testing: A backup is not assumed to exist until it has actually been restored. Our target is to restore a backup to a separate environment at least once a month and verify record counts and that the application starts.

Your data stays yours

Even if you are unhappy with the service, your data should remain with you. Lists export to Excel, reports to PDF and Excel, and documents in file storage can be downloaded. When the contract ends you get time to export, after which data is deleted from the live system. Details are on the data processing page.

Frequently asked questions

How often is the database backed up?

Our target is every 6 hours. In the worst case up to 6 hours of data after the last backup could be affected, which is why our RPO target is 6 hours.

Can you roll back only my company's data to an earlier date?

Backups cover the whole platform. We assess requests to recover specific records of a single company from a backup through support; this is handled case by case, not as a scheduled service.

How do I export my data myself?

Customer, invoice, reconciliation, payroll and report lists can be exported to Excel; files can be downloaded by folder. At the end of the contract we help with a bulk export.

Are RPO and RTO contractually guaranteed?

No. The published values are targets. Contract-specific targets can be discussed on the Enterprise plan.

Related

Let us review our backup setup together

We will share the information your business continuity plan needs in writing.

Write to us