Part 1: The Need for Data Backup and Restore Capabilities
Posted on June 7, 2019 by Steve Pelletier

Journey to the Cloud: From High Availability to Archive

Part 1: The Need for Data Backup and Restore Capabilities

Since the time of the first digital computers the need to keep data safe and available has been a primary concern. In the early days, data backup was simple. Once something was loaded into memory, it could be saved off to tape or punch cards for easy retrieval – easy being a relative term here, particularly with punch cards!

On my first home computer I saved data off to a cassette tape, and eventually a floppy disk which could hold an amazing 140KB of data. I made sure to make duplicates of all of my disks and keep the spare copy in a safe place. Then I got my first 10MB hard drive and I started to get nervous about keeping all of that data in one place. It would take 75 floppy disks to back this hard drive up. That was expensive, and it would be difficult to keep all of those floppy disks organized. What if I made changes to files on the hard drive, how could I back it up again without needing another 75 floppies? Ah, the good old days.

LTO Backup tape for data recovery isolated on white background most use in Data center room.

By the time I entered the world of the employed as an IT professional, companies had servers with tens, hundreds, or even thousands of megabytes of information stored on them, and keeping this data safe was a huge concern. One of my first customer calls as a newly-minted Certified Netware Engineer (CNE) was to a customer that had the single 120MB drive in their server fail. They had been running backups to tape, but had never run restore tests to make sure they could restore files from tape – a critical error. The tapes were stored in a box above their server in a room without climate control, and the tapes were unrecoverable. This is when I realized that companies don’t need backup systems, they need restore systems. And, to get reliable restore systems, they needed multiple copies of their data, a production copy, one for quick restore, and another copy kept off site just in case the onsite backup fails.

Fast forward a few years. I had a customer that had several gigabytes of data stored on multiple servers. They had a solid recovery system (tape backup) implemented, and had contracted with a company to routinely take copies of tapes off-site and keep them in an environmentally controlled vault in the case of an emergency. They routinely ran test restores from their tapes and were very confident that they could withstand server failures. They had a fire that spread to where they housed their servers and lost all of their servers and backup tapes. They thought they were covered because they had reliable tapes safely stored off site. This was true, but they had no place to restore the data. It would take them weeks to get the building cleaned up and get new servers delivered and configured so they could restore their data and resume business as usual. They had all of their data in a safe place, but they needed a true disaster recovery plan that would get them rapid access to a new location and servers to restore their data to. They also needed to take a long look at how frequently they took tapes off site to minimize the amount of data they would lose in a disaster.

Let’s fast forward again, to another customer experience. I had a manufacturing customer that had done their due diligence and developed a very robust data backup and disaster recovery plan. They had identified recovery time objectives (RTOs) and recovery point objectives (RPOs) for all of their servers. They grouped the servers according RPO and RTO requirements so they could identify what types of technologies would be needed to back up their data. They paid fees to a company that would provide them a place and equipment to restore their servers in the event of a disaster, and had worked out how to get their people and data to the recovery site to get the business back up and running as quickly and efficiently as possible. They even did disaster recovery tests twice a year, where they would go to the recovery site and make sure all of their detailed plans worked. The one thing that I noticed in their plans was that recovering their email system was a relatively low priority. I asked them about this and they ensured me that email was more of a convenience than a necessity for them. Several years after they developed the disaster recovery plan, a server failure brought their email system down for nearly a week. Since it was a low priority, non-mission critical system, it did not trigger the criteria for declaring a disaster and invoking the disaster recovery plan. In that week they discovered just how mission critical email was to their organization. They had just discovered the need for high availability (HA) for mission critical applications. Within a month the CEO and CIO had the IT team start identifying which systems were mission critical to the business units, and coming up with plans to make sure these systems never went down.

I have worked with several companies involved in patent disputes. Several of these disputes revolved around which organization came up with the idea for a particular patent first. One way of proving this was for a company to show that they had files related to the idea dated earlier than the other organization. Most organizations were able to retrieve a data backup they had archived from prior years to find the information they needed. Others quickly discovered that the lack of a proper data archival policy prevented them from locating the earliest files they had related to the patent, putting them at a great disadvantage.

In this post, I used experiences from my past to demonstrate the need for data backup and restore capabilities, data archiving, high availability (HA) , and disaster recovery (DR) solutions. My upcoming posts will dive into the technologies used to address these needs, common mistakes organizations make in developing solutions, and how some solutions are blurring the lines between these needs.

Next up, Part 2: Backup and Restore Policies. Until next time, keep your data protected.

Get the FREE eBook

From High Availability to Archive: Enhancing Disaster Recovery, Backup and Archive with the Cloud Ebook

This is part 1 of 10 in the From High Availability to Archive: Enhancing Disaster Recovery, Backup and Archive with the Cloud series. To read them all right now download our free eBook.