Test Everything. Risk Nothing. Upgrade With Confidence.
Note: This is a use case example demonstrating how Rediacc can solve this problem. As a startup, these scenarios represent potential applications rather than completed case studies.
Crisis Scenario: During a database upgrade, an unexpected error occurred that prevented reverting to the old version or proceeding to the new one. Customers couldn’t access systems, and 5000+ employees couldn’t work.
The Problem
Mehmet is an experienced system administrator who manages large-scale databases. He decides to upgrade a 100 TB PostgreSQL database from version 13 to 14. His plan:
- Take a backup → However, backing up takes several days due to data size
- Perform the upgrade on the weekend → Departments are notified of an outage on Saturday 01:00-05:00
Crisis Impact
- An unexpected error occurs during the upgrade
- The database can neither revert to the old version nor proceed to the new version
- Even external support teams cannot solve the problem
Impacts:
- Customers cannot access payment and order systems
- Company employees (5000+ people) cannot work
- Reputation loss and increasing complaints begin
Temporary Solution:
- The last backup is loaded onto a new server → Hardware cost doubles
- Thursday and Friday data are only in the live environment, so data loss occurs
- Two databases with different versions are created → Inconsistencies increase
Rediacc Solution
Mehmet solves the problem fundamentally with Rediacc:
1. Instant Cloning
- A clone of the 100 TB database is created within seconds
- Upgrade tests are performed without affecting the live system
2. Hourly Snapshots
- It is determined which step has been failing since when during the upgrade process
- Problematic operations are identified in advance and corrected
3. Seamless Upgrade
- If the upgrade fails, the live environment is not affected
- If the upgrade succeeds, the new live environment becomes the latest clone
Result
Time and Cost Savings:
- Backup time was reduced from 7 days to 10 seconds
Risk-Free Upgrade:
- Errors were detected in advance in the test environment → No problems in the live system
Zero Downtime:
- Customers and employees felt no disruption