Need Data Off An Old Tape? The Backup Crackers Can Help
March 3, 2025 Alex Woodie
What happens when you need data from an old backup tape, but no longer have the vintage IBM hardware it was backed up from, let alone a subscription to the application it was created from nor the backup software it was created with? If you find yourself in such a predicament, the folks at S2|Data would like to hear from you.
“We are backup software format crackers,” says Brendan Sullivan, the founder and chief executive officer of S2|Data. “What we do is we look at the media, which might be tape or it might be disk, and we figure out how that data was backed up in the first place. And then we run our own software to restore it, process it, reformat it, and present it in a way that is usable for the client.”
That’s it. Or as the kids might say: easy, peasy, lemon squeezy. Except as it so happens, there’s a lot more to the story.
Sullivan’s journey into this particular niche of the digital forensics market began a quarter of a century ago. Around the turn of the century, he was working at Anacomp, which at the time was a tape media manufacturer. TDK, Sony, and Fujifilm were making major investments into tape manufacturing, and the writing was on the wall that Anacomp needed to change its business model, so it spun off a services firm called eMag Solutions.
eMag Solutions morphed into data recovery specialist gradually. Sullivan got a glimpse into what his future would hold back in 2001, when he was asked to help pull data off backup tapes of a bankrupt energy company called Enron.
“We had some software developers and we worked on it and managed to get the data off,” Sullivan tells IT Jungle in an interview. “And we thought, okay, this is pretty cool. And then other projects came along once we got that skill set. And before you know it, you’re sucked into the world of e-discovery.”
Sullivan became president and CEO of eMag Solutions in January 2002, and stayed with the company another eleven years before leaving to join a law firm specializing in e-discovery and tape remediation. In 2013, he founded S2|Data in Atlanta, Georgia to serve the market again.
Unix and Windows account for the vast majority of systems S2|Data works with, but it runs into a fair number of old AS/400s and mainframes. Since many of the company’s tape recovery engagements are triggered by lawsuits and investigations of companies that went under, the big target is often email communications, as opposed to transactional data.
But that’s not always the case. In some situations, a company may simply want to avoid the cost of maintaining a legacy computing environment just for the sake of keeping data for the government-mandated period of time.
“Without a shadow of a doubt, the right way to go about this is to have the AS/400, have the JD Edwards application, and do it that way,” Sullivan says. “But companies go bankrupt, companies retire stuff, companies want to call stuff up eight years after they backed it up. And they just if they don’t have the resources anymore. That’s the time when they would use this approach.”
S2|Data has vintage StorageTek tape drives – and every other drive imaginable.
The process for pulling data off of old backup tapes is similar, even if there are differences in backup software and platforms. There are two main parts.
First, the company must be able to talk to the backup device. S2|Data has access to 30,000 physical tape drives for this purpose, just about every tape drive imaginable. To read the data off the tape, it needs to understand how the backup software wrote the data to the tape.
“It’s understanding, basically, how to run and manage SCSI buffers,” Sullivan says. “And then the far more complex or more varied ones are each backup software, whether it be Veritas or Arcserve or Commvault or TSM or AS/400 BRMS. They all manage data and push it to tape in slightly different ways.”
Once a technician has cracked the code for a particular version and release number of a backup software product, then it creates a DLL for that version and stores it in its Windows application. “So the next time a tape cartridge has been backed up with that same version of software, it’ll say, ‘I know what you are. You’re an IBM label tape version, and I know how to get your data,’” Sullivan says.
Brendan Sullivan is CEO of S2|Data
The next challenge is decoding what the data actually says. Backups in the midrange world work differently than in the Unix and Windows world, and not just the difference between EBCDIC and ASCII character sets.
“Obviously the AS/400 is kind of an object-oriented midrange storage environment. It stores everything, rather than having a tree structure like Windows or Unix,” Sullivan says. “Backups are files and pointers and file marks and that kind of stuff. So it’s taking the data that is on the AS/400 on disk, and it’s simply dumping it with its normal library and object list organization.”
Once the library and object lists are available, S2|Data will run a scan to see what those libraries contain. They’ll look for text descriptors in those libraries that give a hint as to their contents, such as “sales ledger.” Then the company pulls out the raw data, reformat it into something usable, and present it to the customer.
“There is up front energy,” Sullivan says. “JD Edwards might pull in data from all sorts of different objects to create the database that you’re going to read, or the report that you’re going to read. So there is definitely engineering work on the front end to work out how they would have read this or how they would have seen this on the black and green screen.”
S2|Data has worked with thousands of customers over the years. Sullivan was involved in another infamous financial scandal involving an individual named Bernard Madoff, who used an AS/400 to hide his illegal scheme.
While most of Sullivan’s customers involve bankruptcies, criminal investigations, and civil legal matters, S2|Data also has its share of customers who just don’t want to maintain a legacy system anymore. If your data is valuable enough to hold onto, but not valuable enough to justify keeping a vintage AS/400 running on life support, S2|Data could be just what you need.
