Recovery objective
Last updated: Nov 4, 2026
Backups
The database of the Fluid Attacks platform uses DynamoDB Point-in-Time Recovery (PITR). PITR can restore the database to a point in the last 35 days. The most recent point is usually about 5 minutes before the current time. Also, the database has weekly, monthly, and annual full backup schedules. Fluid Attacks keeps the weekly backups for twelve weeks, the monthly backups for three years, and the annual backups for fifteen years.
This table shows the details:
| Type | Date | Hour | Storage |
| Annual | First Monday of January | 6:30 a.m. | 15 years |
| Monthly | Every 1st Monday of the month | 6:00 a.m. | 3 years |
| Weekly | Every Sunday | 5:30 a.m. | 84 days |
| PITR | Continuous | Continuous | 35 days |
Recovery objectives
| Objective | Definition | Time | Description |
| 1. Recovery Time Objective (RTO) | The maximum tolerable time to restore the service to minimum acceptable conditions. If a disaster occurs and the service must be available immediately, the RTO is 0. But, if the organization can tolerate one hour of total service interruption, the RTO is 1 hour. | 15 minutes | The platform is classified as critical, so its backups are continuous. The full restoration of a backup takes 15 minutes, and this is the recovery time. |
| 2. Recovery Point Objective (RPO) | The maximum tolerable data loss in a disaster. For example, if a disaster occurs and the backup of the information is weekly, the recovered service has the data from one week before the incident. Thus, the RPO is 1 week. If the backup is online and no data loss is acceptable, the RPO is 0. | About 5 minutes | DynamoDB Point-in-Time Recovery (PITR) can restore the database to a point about 5 minutes before the incident. |
| 3. Maximum Tolerable Period of Disruption (MTPD) | The maximum time before the service must operate normally again to prevent losses that the organization cannot accept | 1 month | The monthly operating costs and the monthly revenues are almost the same. Thus, if the service stops completely for more than 1 month, the losses can be unsustainable. |
Impact over time
A service interruption can occur at any time. But, at some times, the interruption can be more severe than at other times. The sections below analyze this.
Monthly analysis
In the first and last months of the year, the use of the service decreases because of platform freezes and customer vacations. Thus, the impact is less severe than in other months.
| Month | Impact |
| January | Medium |
| February | High |
| March | High |
| April | High |
| May | High |
| June | High |
| July | High |
| August | High |
| September | High |
| October | High |
| November | High |
| December | Medium |
Weekly analysis
In most cases, projects start or close in the first and last week of the month. Thus, the demand for the service can increase, and the impact is higher than in other weeks.
| Week | Impact |
| 1 | High |
| 2 | Medium |
| 3 | Medium |
| 4 | High |
Daily analysis
On all the days of the week, disturbances can have a high impact. But most customers do not work on weekends. Thus, the impact is lower on those days.
| Day | Impact |
| Monday | High |
| Tuesday | High |
| Wednesday | High |
| Thursday | High |
| Friday | High |
| Saturday | Low |
| Sunday | Low |
Hour analysis
Fluid Attacks estimates that disturbances in the business hours of the customer have a higher impact than issues at other times of the day.
| Hour | Impact |
| 0 | Low |
| 1 | Low |
| 2 | Low |
| 3 | Low |
| 4 | Low |
| 5 | Low |
| 6 | Low |
| 7 | High |
| 8 | High |
| 9 | High |
| 10 | High |
| 11 | High |
| 12 | Medium |
| 13 | Medium |
| 14 | High |
| 15 | High |
| 16 | High |
| 17 | High |
| 18 | High |
| 19 | High |
| 20 | Medium |
| 21 | Medium |
| 22 | Low |
| 23 | Low |