Sep 13, 2021 | Paul Reeve
We are all aware of cloud storage as a means of storing our important data. Cloud storage can be used for backups, archives or even as an extension of local file systems for current data. Many cloud providers now offer different storage tiers and if you want to store and manage data cost-effectively, it’s important to understand how these tiers work.
This article will provide an overview of what these tiers are and their application usage. As costs can vary from region to region, we won’t be looking at a precise cost breakdown. However, we will be including resources where you can find more information about pricing by different cloud providers.
Cloud tiers are broadly categorised as Hot or Cold. Hot tiers are used to store frequently accessed data. The storage costs will be higher, but data access is immediate and access charges for data will be lower or free. There will also be no minimum contract lengths for data storage. Cold tiers are for infrequently accessed data. Storage costs will be much lower but there will be minimum contract lengths. Data is generally not immediately available and can take several hours to retrieve. There will be significantly higher charges for retrieving data.
Not all cloud providers offer tiers. Some offer flat rate single tiers which need to be considered for certain types of data workflows. Pricing here is typically lower than other hot storage but higher than cold. Data is immediately accessible and there are usually no access charges. There may be minimum contract lengths for best pricing. A big advantage is that monthly bills will not vary if workflows change unexpectedly.
AWS currently have six tiers – four hot tiers and 2 cold tiers, which are all S3 compatible and use the same API. They are listed below in order from hot to cold.
S3 Standard, as the name suggests, is the primary hot tier. It is the most expensive storage tier, typically priced at around $20 per TB/month, but data is immediately accessible and there are no data access charges or minimum contract lengths. This is an ideal tier for local system file extension.
S3 Intelligent – Tiering is used for workflows with unpredictable changing patterns. It will automatically move data between various hot and cold tiers based on usage. The appropriate tier storage charges apply and there is a management charge and a minimum object size charge. It does make for very unpredictable costs but if there is no other way to automatically manage the data lifecycle, this may be a good choice.
S3 Standard – Infrequent Access: this is a lower-cost hot storage tier for infrequently accessed data that still needs immediate access. It is around half the cost of S3 Standard but there is a minimum object size charge and a 30-day contract for each object stored. Access charges are also applied. This is a good option for continuous data protection and file extension where data needs to be immediately accessible but not frequently.
S3 One Zone – Infrequent Access: One Zone is specifically designed for backups where this is one of two physical copies of the data. It is the lowest-cost hot storage tier at around $10 per TB but data is only stored in a single data centre in a region, which means there is a higher risk of data loss.
S3 Glacier and Deep Glacier are the two cold tiers, designed for archive data that rarely needs to be accessed. Storage costs are $4 and $1 per TB/month, respectively, but data is only accessible as standard in 12 hours. Glacier offers expedited retrieval in minutes but at a much higher cost. Both tiers have higher data retrieval costs and minimum contract lengths of 3 and 6 months per object. Many organisations that need to store data for compliance reasons would do well to consider these tiers as it will dramatically reduce long term storage costs.
Microsoft offer 3 tiers: Hot, Cool and Archive. There is also a premium hot tier which is designed for high transactional data, but it is not currently possible to tier from the premium tier to other tiers. Otherwise, the Hot, Cool and Archive tiers follow the same pattern of high-to-low storage costs, low-to-high data retrieval costs and data retrieval times as AWS S3 tiers. The minimum object storage contract periods are zero, 1 month and 6 months for Hot, Cool and Archive, respectively.
Microsoft do offer reserved capacity prices which provide a lower fixed price per tier for a minimum storage commitment over 1 or 3 years.
Google offers 4 tiers: Standard, Nearline, Coldline and Archive. Pricing does vary across regions, but it is broadly in line with AWS Standard, Infrequent Access and Glacier. There are no data retrieval costs for Google Cloud Standard, but they are higher across the tiers. It is worth noting that the Archive tier has a minimum 365-day contract period per object. On the upside, Google Cloud does provide immediate access to archive data if required.
IBM also have 4 standard Tiers. Like Amazon S3 Intelligent Tiering, the IBM Smart Tier will manage your data between hot, cool and cold based on activity and access. Then there is Standard, Vault and Cold Vault for different data retention classes. Again, storage prices range from high to low and data retrieval times from low to high. They also have Archive and Accelerated Archive classes which offer 12-hour or 2-hour retrieval times with lower long-term storage charges.
Wasabi are a new generation cloud storage provider offering only cloud storage at a fixed price of $5.99 per TB per month (this may vary by region) with no ingress, egress or I/O costs. Some contracts may have a minimum object storage charge of up to 3 months. The cost is also lower for reserved capacity contracts for over 1, 3 or 5 years.
Other cloud providers like Seagate, RSTOR and Backblaze have a very similar model.
While there is a huge amount of choice of cloud providers, this is often narrowed down by company policy, which regions are serviced by which providers or data sovereignty concerns.
There are also other key data management criteria like versioning support or bucket immutability which can be factors in the decision-making process.
Your cloud choice may also be influenced by other cloud services such as compute services. Azure and AWS have multiple services in addition to cloud storage.
That being said, the key to determining which cloud services to use is understanding your data and workflow. Tiger Technology offers a storage assessment service which helps you understand the data you have, how and when it’s used and therefore which cloud storage service tiers you should consider.
The following broad principles can be applied depending on your primary business needs & data management goals:
Tiger Bridge is a policy-based data management software-only solution that has full API integration with all leading cloud providers. Because Tiger Bridge understands all the metadata associated with your local data, you can set policies based on data age or data usage to copy or move data to any tier in the cloud. This means that data can live on the lowest-cost tier according to its data lifecycle.
Tiger Bridge replaces the intelligent data management cloud services which have relatively high management charges and automatically moves data between tiers based on policies. Tiger Bridge’s flexibility means you can also move data directly to archive tiers if, for example, it’s designated compliance data.
A single Tiger Bridge instance also supports multiple cloud targets giving businesses the complete freedom to have their data stored cost-effectively.
Cloud tiers offer huge flexibility for storing your data cost-effectively. There are lots of options and pricing is often very complicated to determine due to the array of storage, access and management charges that can be applied. Cold storage is very cost-effective but there will be minimum contract lengths for objects, generally high egress charges and time delays for data retrieval.
Understanding your data lifecycle is key to making the right decisions. If you do not have the tools to obtain this information, consider a data assessment service which will show past and current data usage patterns. Once you have this information, consider what data management tools are available. Tools that have full cloud API integration like Tiger Bridge allow you to automate this process easily on an ongoing basis rather than having to do this manually and ensure you avoid making costly errors. Always monitor your cloud costs so you can adjust your policies if data patterns change.
Want to learn more about using Tiger Bridge to connect your local storage with cloud tiers? Check out this 5-minute overview video.
Ready to try it out for yourself? View our Tiger Bridge subscription plans.