It’s time for the final installment in my three-part series on cloud/online backup FAQs. In the first two parts I provided a general overview of cloud and online backup deployments, addressed concerns around data security, and offered details on the benefits of cloud/online backup in the event of a disaster.
To wrap things up, let’s look at more technical details of data backup processes.
1. When and how often do data backups run?
This depends on the method being used, your end goals, and what exactly you are backing up. For example, a domain controller doesn’t need to be backed up often; backups for these might only run once per week. However, something like a file or exchange server is changing all the time, so to give you the best granularity for restoration points, you may want to back these up hourly.
Backups run depending on restoration points or recovery point objectives. A standard for backups is typically every 24 hours. Increasing or decreasing frequency should be driven by this question: “How much ‘processed business’ (or number of business transactions) can I afford to lose?”
If your business is an online catalogue ordering platform, you’ll stop processing business the minute your data is no longer connected to your systems. As such, you may want backups to run more frequently than once per day. This would provide what we call rollback or restore points throughout the day. Conversely, if your business processes orders weekly in a “batch mode,” you would most likely be covered by weekly backups completed just prior to the weekly run.
2. How long does it take to back up my data?
The first time you back up your production data, the process will likely take a long time. After all, you’ll likely be backing up many gigabytes or terabytes of data at once. While the initial backup will be lengthy, your ongoing backups will happen much more quickly as these incremental backups need to only replicate changes in your data since the last backup point.
3. How long are backup copies available?
Service providers will typically keep backup copies from the most recent restoration point(s). This should, by nature of the most recent backup having contained all data that came before it, suffice to restore your production environment.
Service providers may also set systems to save backups for a few days prior to each recovery point, so that in the case of a recovery situation there are multiple distinct recovery points. This is important because, unfortunately, servers don’t always come back into production perfectly.
Service providers should always support your business’ legal requirements for retention. Ensure that your data is maintained for your specified length of time. The use of tiering is also applicable here to minimize the cost of long-term online and cloud-based storage. A hybrid backup approach, combining tape and online models, is often used for keeping some data close at hand and other data in long-term “legal retention” mode.
4. Are there bandwidth optimization tools to prevent backups from interrupting other applications and business processes?
Yes, there are, but the best practice here is to schedule your backup cycles at times that will not interrupt critical processes. Often, data backups will occur later at night so as not to disrupt work during usual business hours. Work with your service provider to define the timing and processes to best suit your needs.
5. How will tape vs. online cloud backup affect my team’s productivity?
On a day-to-day basis, online backup (or, for that matter, any other backup method) should not at all affect your office’s productivity. It should be transparent to you as a user. In the event of a disaster and recovery scenario, however, cloud backup will help restore your team to normal productivity levels as soon as possible by providing access to data and applications.
Keep in mind that your choice of backup model will, however, affect your team’s productivity in a post-disaster scenario. When restoring your data, it will take significantly longer to restore your data to your team due to tape transit and load times. With online and cloud backup, your service provider can restore data at a much faster pace and thus allow your team to focus on higher-level business restoration activities.
6. What determines the price of backup for tape, online and cloud-based models?
Pricing is usually dependent on factors like the amount of data and the type of disk used in the backup process. There are some software elements in each of these solutions for accessing and writing the data. Don’t forget about the physical equipment needed if pursuing a tape solution (readers, spoolers, tape libraries) and tier of data. You will need to discuss all of these elements with your service provider at the outset to ensure that the solution fits both your needs and your expected price point.
7. Where is my data actually stored?
When implementing an online backup model, your production data resides (and is used) within your office while backups are copied to, and stored in, your service provider’s data center(s). In the hybrid approach suggested earlier, a backup of your data is also stored locally in your office on disk so recent backups are readily accessible without having to download huge backup files from your cloud.
When determining the location of your backup(s) – that is, the location of your service provider’s data center(s) – consider geographical redundancy to better prepare you for a disaster. If your data center is in the same city as your office, a localized disaster will still leave you unable to work since both your production and backups will be inaccessible. Instead, it is critical to have at least one instance of your backup files in a completely different geographic zone. Xtium, for example, backs up customer data to centers in Pennsylvania and Nevada.
Remember, having a functioning and disaster-ready continuity plan is vital for your business. Whether you choose cloud/online backup or any other method, ensuring that your plan is in place and tested can be the difference between your business reopening soon after a disaster, or never reopening at all.
Meet the Author: Shawn Fichter is a data center veteran who has been working in the DR and business continuity space for over 15 years. Over the course of his career, Shawn has been involved in managing infrastructures through the transition of client server to utility computing to cloud-based service models. He now leads the development of next generation cloud-based products and services for Xtium.