How to do IT: Backup Planning

If this server blows up, how do you recover the informaiton?

This is part of an ongoing series in the publication HowITsDone

One of the most important things you can do is ensure you have a copy of your digital life somewhere. Think about it, in today’s world, everything is digital and because of that there is an inherent risk that a failure or a mistake can wipe out entire swaths of important things in an instant. What would happen right now if you computer or phone just blew up? I am not talking failed and repaired, I am talking about a stick of dynamite taped to your device, lit and then thrown into the ocean. Could you recover everything? Can you recover anything? That is what we are going to talk about today, starting with an IT professional position and in another part the everyday person.

From an IT point of view, whether your company is small or large, you more than likely have a backup policy in place, enforced or not, that dictate what you do and do not backup. If the information has been identified properly, then backing up becomes a question of how to get the process standardized, frequent and reliable. There are a ton of solutions out there to choose from and countless ways to approach backing up that makes sense to your company. Here are a few things to keep in mind for your business when building a backup strategy;

Strategy Keys

  • Discovery- What information is important enough that you want to keep it after changes have been made or over time? Or better, what is needed to keep your business running and couldn’t be replaced at all if it was removed?
  • Retention- How long back in time would you potentially need to go to recover information if it is lost? What are you willing to stop at? If attempting to keep long term data, how much will be retained on site, versus off site storage? Also how many copies of something do you want to keep?
  • Recovery- How quickly does information need to be restored. This is normally broken down based on Retention into tiers, for example, any information from the last 30 days needs to be able to be recovered within an hour, anything past that a couple of days or etc.
  • Testing- Always and forever test your backups. Depending on the size of the environment, it might be possible to do a test restore of every weekly backup. However, performing at least one random restore every quarter will help in the confidence of what is being backed up is actually being backed up.

Above are some of the core, high level principals of data protection and each has their own nuances to them, but for the most part answering the above fleshes out the majority of questions and will get you far enough that you can start actually looking at solutions and policies to work around the specific needs of the business. Knowing the expectations of what the company has in mind when a restore is needed is critical. This sets up everything and allows the executives, managers and administrators know exactly what is expected and how things are being done in the event of a recovery.

Once the workflow has been established, the course of action is to find the solution that meets all of your criteria. It is important to define this beforehand because there are plenty of options out there and you may get side tracked on the bells and whistles of all of the different ways data can be protected. Just follow your guidelines and it will make whittling down the different products that much easier. The more automated you can make something, the better for the most part. It just frees the back up administrator to do more things and makes everyone more productive.

There are many ways to solving this problem. I have seen everything from keeping things on redundant external drives and somewhere in the cloud, to complete enterprise solutions that not only allow for complete disaster recovery, complete with copies on two different company sites, with monthly tapes going off site while additional copies are kept on site. It just comes down to what your company needs are and how to support those needs.

Having at least two additional copies of information besides for the live data is ideal, especially for recent data. Keeping a well defined set of tiers will help keep different levels of granularity throughout your backup environment. It also helps with keeping things accessible. No one wants to have to travel to a cave every time someone deletes an excel sheet.

Types of Backup

Depending on the size of your environment and the amount of change that occurs, most backup scenarios incorporate a type of partial data backup. Here are the some of the basic kinds of protection processes out there.

  • Full- Simply a complete backup of all relevantly identified data.
  • Differential- This backup does all of the changes that have occurred since the last backup. What this means is that if a Full occurred on Sunday, and Monday there was 1 GB of change, then it will backup that changes. By Wednesday however, if there were 1 GB on Monday, 10 GB on Tuesday and 3 GB for Wednesday, then that nightly backup will consist of all changes and back them all up at that one time.
  • Incremental- Like a differential, however only the changes since the last backup remain. So only the changes that were made since the last backup of any kind are taken
  • Transaction Log Backups- Primarily for databases, these backup the transaction history periodically on databases to allow for more granularity on things that change frequently

There are other types of backups that are a little more complex, but for the purposes of this article, pay attention to these above. With your plan established, take into consideration the size of your data that needs to be backed up. Best case scenario is to take a Full backup everyday, but it often isn’t feasible, but daily changes is something that should be targeted. This will allow you to have daily changes in the short term, depending on established needs from your plan.

That leaves Execution. Ensure that they get done and are working. Check on them because even if there aren’t failures, a bad configuration could keep something from backing up and you will never know that it isn’t working until it is time to recover it. Think of this as protection, a way to ensure that the information your company needs is all ways available. For the IT professional, backups are one of the most important aspects of their jobs, yet tend to be one that is often thought about last, do not let a restore failure derail your business.

This information should help individuals and businesses alike become more efficient and have proper discussions about how their backups are going and should be going. It is a process, that once established, should run fairly seamlessly, especially when the proper checks are in place.

If you liked this please recommend it and follow along at HowITsDone. Let me know if there is something else that you would have included or if this kind of information helps you out. I help businesses with technology and developing workflows to help them get more done. Find me on twitter @hap1o or shoot me an email