Backup and disaster recovery are typically cited among the top use cases for cloud computing. Companies are increasingly realizing that cloud-based backup is more efficient, more reliable, easier to use and less error prone than traditional methods such as tape or other rotational media, not to mention it enables extremely effective disaster recovery. In 2015, we saw a major upswing in cloud adoption for storing backup copies. In 2016, expect to see cloud backup and disaster-recovery-as-a-service (DRaaS) options become the norm for new backup deployments at organizations of all sizes.
Most companies can easily lower OPEX and CAPEX with cloud backup. But not all cloud backup options are the same, and a true cost comparison entails evaluating more than just each offering’s top line cost per GB. In this slideshow, cloud-recovery leader Unitrends details five “hidden” charges that companies must consider when researching the best and most cost-effective cloud backup solution for their business.
Cloud Backup’s Hidden Charges
Click through for five “hidden” charges that companies must consider when researching the best and most cost-effective cloud backup solution for their business, as identified by Unitrends.
Compounding Storage Needs
Hidden Cost No. 1: Compounding Storage Needs
Retaining backup copies is now a necessity for many companies for a variety of reasons, from protection against human error, natural disasters and system failures, to compliance with regulatory or policy requirements.
The storage required to retain multiple copies of your backups can be substantially larger than the original source data. A common strategy for keeping data for long-term retention is the grandfather-father-son (GFS) backup scheme, which results in a lot of backup copies. In this scheme, a set of daily, weekly, monthly and yearly backups are retained using a first-in-first-out (FIFO) pattern. Even if only the data that is changing daily is stored in the cloud and the backups are being compressed, the amount of cloud storage used can add up over time.
Consequently, it matters a great deal where the gigabytes (GBs) and terabytes (TBs) you are purchasing reside. Most cloud storage and backup solutions charge for every GB of storage used in the cloud. So, as your backups grow and consume more cloud storage, so does your bill. But there are cloud backup solutions that only charge for the amount of data protected on the source side. Evaluate each cloud backup solution to determine whether you’ll be paying for protected data or raw cloud storage.
Retrieval Fees
Hidden Cost No. 2: Retrieval Fees
Many of the low-cost hyperscale cloud storage solutions are designed as cold storage, meaning they are intended for storing data that does not need to be frequently accessed. These cold storage solutions charge a retrieval fee to users who need to access their data. Though these fees are typically 1 cent per GB, they can quickly add up if you need to access your data frequently throughout the year.
Network Egress Fees
Hidden Cost No. 3: Network Egress Fees
All clouds allow you to insert your data for free (network ingress). However, many clouds charge network fees to move your data out of the cloud to another location or to the Internet (network egress). These network egress fees vary greatly by cloud provider and depend on the amount of data moved in a month as well as the geography where the data resides, though they tend to range between 5 and 23 cents per GB.
Getting Data Into and Out of the Cloud
Hidden Cost No. 4: Getting Data Into and Out of the Cloud
Getting your data into and out of the cloud has both hard costs and soft costs beyond network egress fees. The impact of those costs depends on which option you use to move your data – a WAN connection or physical seeding. The most obvious method to move data to and from the cloud is to send it over a WAN connection. However, this may not be ideal for moving large amounts of data because, even with a high-speed Internet connection, transferring multiple TBs of data can take a substantial amount of time.
Physical seeding is an alternative method to get an initial set of data into the cloud without having to wait days or weeks to transfer many TBs. Physical seeding uses physical disks and overnight shipping to quickly create an initial dataset in the cloud. Media is sent to the cloud provider to “seed” the initial full set of data and avoid the WAN challenge.
Physical seeding can greatly speed up the cloud backup initialization process, reducing setup time from weeks to a single day, even for hundreds of TBs. However, not all vendors offer a seeding option and almost all charge a nominal fee for it.
Getting Data Into and Out of the Cloud
Hidden Cost No. 4: Getting Data Into and Out of the Cloud (Continued)
Even more important than seeding your data to get started is the requirement for “reverse seeding,” or a data shipment service level agreement (SLA), to get data back within 24 hours in the event of a disaster. This works the same as seeding but in reverse. If you have a disaster and lose all or a large amount of your data, the cloud vendor places your data onto disks or a new backup appliance and ships the data to you.
Whether an organization can obtain reverse seeding or not can be the difference between achieving or missing recovery time objectives (RTOs). With reverse seeding, you avoid the challenge of sending large amounts of data over a WAN, and downtime can be minimized from weeks to hours.
Cost of Not Having a Solid SLA
Hidden Cost No. 5: The Price of Not Having Rock Solid SLAs
Many clouds have soft SLAs. It pays to read the fine print and determine how the cloud you are relying on for mission-critical backup and recovery will perform when it matters most. Consider the following questions when evaluating cloud providers:
- What does the cloud provider guarantee for recovery? Does the provider guarantee a specific recovery time?
- What is the guarantee for uptime?
- If you have concerns about data sovereignty, does the provider validate that your data will stay in the geography that you choose?
- What are the SLAs for response time for service? Is there a person to answer the phone or is it email/chat only?
- Are there multiple vendors involved? Does that affect the SLAs?
SLAs are particularly important for backup and recovery. You need to make sure that you have a supplier who will be there with you when you have a disaster or emergency. Without such SLAs in place, you may find that you have just purchased cheap offsite storage rather than a true backup and recovery partner.
