Cloud storage costs: How to get cloud storage bills under control

Key cost elements

The cost of enterprise cloud storage can vary greatly dependent on particular use cases. Accessing data, capacity stored, retrieval frequency and network bandwidth all add to the cost.

Capacity

One of the key cost elements is capacity. If a large enterprise wants to leverage the benefits of the cloud, it can mean migrating and hosting petabytes of data, and this comes with a considerable cost implication, according to Dave Locke, Europe, Middle East and Africa (EMEA) chief technology adviser at independent technology service provider WWT.

“It is a simple problem of numbers, but these numbers are extremely large. Businesses are asking themselves, ‘How do we manage large datasets in a cost-effective way?’,” he says.

Not only can the datasets be extremely large, but there is often a multiplying factor to cost when capacity is considered, he adds.

“If a global business utilises native cloud services, it will have to host a copy of its data across multiple locations, such as America, Europe, Asia and so on. There are costs associated with every copy of this data. But the costs don’t end there,” says Locke.

Operations

There will also be costs associated with actions taken on data in cloud storage. A good example of this is listing information about buckets and objects in Google Cloud, says Justin Augat, vice-president of marketing at Iland.

“Operations are usually priced based on thousands – or tens of thousands – carried out, and can be further broken down by classes of operations which can be priced differently – class A, B, C, and so on.

“Not all vendors classify operation costs the same, so make sure your use case is optimised for operations cost. In other words, if your cloud storage is used for static data such as a long-term backup, you would want to prioritise the cost of storage capacity over the cost of operations against the storage,” he says.

Networking and egress costs

Cloud storage is not just about how much it costs per gigabyte (GB) stored. There are also costs associated with the transfer of data in and out of the cloud.

In many services there are two costs: one per-gigabyte cost each time servers in different domains communicate with each other, and another per-gigabyte cost to transfer data over the internet.

“For example, in AWS [Amazon Web Services], you are charged if you use a public IP address. Because you don’t buy dedicated bandwidth, there is an additional data transfer charge against each IP address – which can be a problem if you create websites and encourage people to download videos,” says Richard Blanford, CEO of IT service provider Fordway.

“Every time a video is played you incur a charge, which will soon add up if several thousand people download your 100MB video.”

Augat warns that, in some cases, providers also charge for Egress bandwidth. “This is an additional network activity cost that is usually assigned if the user is outside the region of the cloud storage,” he says.

Disaster recovery costs

The same issue applies with resilience and service recovery, where you will be charged for data traffic between domains to keep a second disaster recovery (DR) or failover environment in a different region or availability zone. 

“Moving data between regions and out of the public cloud also incurs a fee. Most companies that use a public cloud service pay this for day-to-day transactions, such as moving data from cloud-based storage to on-premise storage, and costs can quickly spiral as your tenancy grows,” says Blanford.

Cloud storage charges of the big three

Cloud providers really don’t make the job of comparing costs an easy task for enterprises. In fact, a whole industry has built up around the use of third-party cost management that aim to help gain a clearer picture of cloud costs.

For AWS, for example, the number of variables make it hard to obtain accurate estimates of costs. Microsoft is little better as its complex software licensing options, discounting and pricing structure can be difficult to understand without considerable knowledge of the subject. Google is perhaps a little easier as pricing is a point of differentiation.

For AWS, its cloud services pricing guide can be used to find out what its latest prices are, and the same goes for Azure and GCP.

Below is a quick comparison of the prices charged in the “hot” object storage class for frequently accessed data by Amazon S3, Microsoft Azure Hot Blob Storage, and Google Cloud Storage standard.

Cutting the cost of storage

Nowadays, most cloud suppliers offer various tiers based on the speed of each type of storage. Putting data in the wrong tier can soon increase costs way more than necessary. It is therefore important to figure out how much data needs to be stored in the cloud, how much you will access it and how important it is to your business.

Cloud providers price storage based on how much storage is used, where it is stored and the type of service. Enterprises can choose from online, nearline, offline, archive and other storage types, so choosing the right one will be more cost-effective.

Some storage services bill by the month, so if you have to store data for only a few days, another service with no minimum billable period may be cheaper.

For cheaper storage, thin provisioning is key. It means you can ensure your organisation is billed for capacity used, not the total provisioned.

Enterprises should also use the monitoring tools offered by cloud providers to see public cloud usage and expenses so they can cut excess usage and remove instances that are no longer in use.

Unnecessary storage can increase costs, so remember to delete data that is no longer in use. Remember also that some archive services can charge for early deletion.

Finally, cut down on the amount of data movement. Public cloud storage should not be used as a primary backup location. Moving large amount of data from the cloud can result in big costs. Data deduplication can also decrease the storage capacity you need. Also ensure that applications are designed to minimise data transfers to and from the cloud to decrease cloud storage costs.