Legacy Server Migration & Database Recovery

The problem: An emergency call. A client in the professional services sector contacted us in the morning with a situation they had feared for months: the primary server had partially failed overnight. The operating system was Windows Server 2008, unsupported since 2020, and the SQL database was partially corrupted after a hardware failure at the RAID array disk level.

The situation was made worse by the fact that the backup policy had not been followed. The last valid full backup was from several months earlier. Potentially lost data included contracts, client history and financial records. The panic was justified.

How we approached it: We arrived on-site the same day. The first thing we did was stop any writes to the disks to prevent overwriting recoverable data. We created bit-by-bit images of the affected disks before any intervention — an essential safety measure that many skip.

Forensic analysis showed that the database structure was partially intact. We used specialised SQL page-level recovery tools to extract critical tables one by one, validating the integrity of each. The process took almost 14 hours, but it was worth it — we recovered 100% of the critical data.

Simultaneously, our team prepared the new infrastructure: physical server with Windows Server 2022, updated SQL Server, RAID 1+0 configuration for redundancy, and an automatic backup policy with Veeam — daily local backup, weekly cloud backup, with monthly scheduled restore tests.

The actual migration was done over the weekend, with zero activity from the client. On Monday morning, employees logged into the new system without knowing that behind it was a completely different infrastructure.

Results:

• 100% critical data recovered, including data considered lost
• 0 hours of operational downtime perceived by users
• Infrastructure modernised by 14 years compared to what existed before
• Automatic backup policy implemented and documented
• First restore test performed in the client's presence — they saw with their own eyes that it works
• The client no longer depends on a single point of failure