A Practical Guide To Data Recovery

Chongwei Chen is the President & CEO of DataNumen, a global data recovery leader with solutions trusted by Fortune 500 companies worldwide.

As a data-recovery expert with 24 years of experience, I have witnessed countless examples of companies facing catastrophic consequences when faced with data loss.

Take, for example, a mid-sized manufacturing company I worked with that could not access its production database due to a hardware failure. Although they had regular backups, the latest incremental backup file was also corrupted. Because of these issues, they had to pause production for several weeks, causing losses of about $1.2 million.

Unfortunately, this company’s experience isn’t unique. A 2022 Arcserve study found that 76% of businesses lost mission-critical company data. Verizon research supports this, concluding that small instances of data loss cost businesses between $18,000 to $36,000, while large-scale incidents can cost up to $15.6 million. The stakes are so high that, according to the University of Texas, 94% of companies facing catastrophic data loss don’t survive—43% never reopen, and 51% shut down within two years.

Given these risks, understanding how to recover data is critical. Let’s look at common storage methods and recovery techniques that organizations should be familiar with.

Data Storage Methods

In modern computers, data is generally stored logically as files, which are managed by a file system. Companies typically use two types of infrastructures to store data:

• On-Premises: This includes traditional hard drives, USB flash drives, SD cards, CDs, DVDs, etc.

• Cloud: Today, over 60% of all corporate data is stored in the cloud, according to G2 research, which includes Google Drive, Amazon S3 Storage, Microsoft OneDrive and so on.

While human error is the leading cause of data loss, other causes include hardware failure, theft, software corruption, viruses, natural disasters and power failure.

Classifications Of Data Recovery

Data recovery is closely linked to the storage methods used to preserve the data, and the recovery techniques can generally be classified into these two categories:

1. Hardware-Level (Physical) Recovery

This method is geared toward hardware failures in storage devices, and it focuses on using the most advanced hardware technologies to:

• Replace damaged interfaces, circuit boards or write heads.

• Use specialized devices or environments to extract data.

2. Software Level (Logical) Recovery

This recovery method uses an advanced software algorithm. There are two sub-categories:

• Raw-Level Recovery: Generally deployed when the target files are lost due to issues like accidental deletion or reformatting the disk by mistake, the data-recovery software scans the raw disk or drive and recovers the files. In general, this software will support multiple file types.

• File-Level Recovery: This method is used when target files exist but cannot be opened by the necessary application due to file corruption. Normally, for each file format, there will be a dedicated tool from the designer of the file format to check the integrity of the file and fix errors in it. For example, for an Outlook PST file, Microsoft provides an Inbox Repair Tool (scanpst.exe) that can scan and fix errors in the PST file. For a SQL Server database file, a SQL command—DBCC CHECKDB—can check the integrity of a database and fix it if necessary.

Why Multiple Techniques Are Often Necessary

The two above classifications are not absolute. In real-world practice, multiple techniques may be required.

For example, consider a situation where a criminal deleted a database containing financial data from a hard drive and then used software to overwrite the entire hard drive. Start by using a hardware method to recover most of the raw data from the drive. Then, apply raw-level recovery software to scan and extract the database file. If the recovered file still isn’t recognized by SQL Server, use the DBCC CHECKDB command to attempt a repair, hopefully recovering most of the financial records from the database.

These techniques can also be very flexible, and the techniques in one category can often be applied to another category to obtain better recovery results or lower the cost.

For example, some file-level recovery software can also recover data from the hard drive directly if no files are available, which will normally offer a better recovery rate than using a raw-level recovery tool first and then a file-level recovery tool second. Some raw-level recovery software can also recover files with hardware issues, such as bad sectors, which will lower the cost because this method does not require specialized hardware devices.

Conclusion

Data loss is often unavoidable, but it doesn’t have to be a disaster if organizations familiarize themselves with the proper planning and techniques. To minimize the impact of data loss and ensure a swift recovery, organizations should follow a few essential best practices:

• Prevention is the most important. Design a comprehensive business continuity plan, including a regular backup strategy and an incident response plan. Implement this plan strictly.

• Act quickly after a disaster. Once you know there has been data loss, respond quickly, ideally within 48 hours of the incident.

• Get professional help. For complex cases, seek professional help from data-recovery experts; they have likely seen the issues you’re facing before, and they can help design the best recovery strategy.

• Implement post-recovery review. After the incident, update the continuity plan and backup strategy based on the findings to reduce the likelihood of future incidents.

With these best practices and by responding strategically, companies can often turn a data-loss incident from a crisis into a manageable challenge.

Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?