As a SaaS Data Intelligence company, Collibra is responsible for backing up your data and then restoring that data, when required, in response to various types of incidents. The following describes our backup and recovery practices and timelines for recovery from different incident types.
Repository backups
We perform backups of databases and file systems according to the practices below. With these, we can restore customer data for the last 30 days with an accuracy window of 15 minutes.
| Category | Backup practice |
|---|---|
| Database |
We support point in time recovery with an accuracy window of 15 minutes for the past 30 days using a combination of full and incremental backups. This system is fully separated from the Collibra Console backup and restore feature. |
| File systems | We take a nightly snapshot of all disk volumes, except for the root partition. The snapshot is taken according to the time zone of the server. |
For disaster recovery purposes, we ensure that backups are replicated across multiple availability zones located in the same cloud region. These backups are encrypted at rest using AES-256 encryption.
We can restore your system or data as required throughout the subscription term. At the end of the subscription term, the final backup is retained and available for 30 days after subscription termination.
Note Currently, we don't support multi-region backups.
Recovery
The procedure to recover an encrypted backup depends on the event that triggers the need for recovery.
| Incident type | Recovery practice |
|---|---|
| Data loss | We restore data to the last known good state, within 15 minutes of accuracy. |
| Database corruption | We analyze the database dump to find the time stamp of corruption. After this investigation, we restore the database to the last known point in time prior to the corruption. |
| Server problems | We analyze server problems. If the problem is not found in a reasonable time, then we schedule a full restore, but only with your approval. The application and the data can be restored within up to 8 business hours. |
| Availability zone crash and data loss | We restore an off-site backup in a different zone. |
Note We only offer hot/cold recovery across multiple availability zones within the region.
About RPO and RTO
RPO or Recovery Point Objective, is the time from the last data backup until an incident occurred that may have caused data loss.
RTO or Recovery Time Objective, is the time that you set to recover the lost data.
Given the backup and recovery practices, we support the following RPO and RTO:
| RPO | 30 days with a granularity of 15 minutes. |
| RTO | Up to 8 hours under normal circumstances, depending on the incident type and the volume of the data to be restored. |