Saving on AWS storage with newly taggable Glacier vaults
If tags are an AWS Cost Management guru’s best friend, then the opportunity to pay a penny per GB per month on archival storage is a close second. Which is why we’re excited that this week, Amazon Web Services announced that you can tag your Amazon Glacier vaults for more comprehensive cost allocation.
In light of this announcement, Amazon Glacier has become an even more powerful tool for minimizing AWS costs than ever; but even the most effective tool will only work if it’s wielded right-side-up. Here are some strategies to keep in mind for saving with Glacier.
Allocating costs with tags
As with any AWS resource, ensuring that your Glacier vaults are tagged according to how they’re being used will be key to understanding why your Glacier-driven costs are what they are—and whether you’re spending efficiently.
You can assign up to 10 tags to each of your Glacier vaults according to any number of categories, including department, owner, role, or name, in order to allocate your costs and see how your Glacier storage spending breaks down by those metrics. Locating your biggest sources of Glacier spending and determining the value of keeping those vaults around (do you really need to keep the data from that old project, in spite of having never accessed it since it was archived?) can represent a considerable source of savings right off the bat, and monitoring your spending according to these metrics on an ongoing basis can ensure that your vaults continue to be used efficiently.
Archiving from S3
Since Glacier storage costs considerably less than S3, archiving data from S3 to Glacier can save you big; you just have to know how to identify candidates for moving over.
Keep an eye out for S3 buckets that aren’t being accessed; if these buckets aren’t receiving “GET” requests, you have little use for the online storage and rapid retrieval that S3 costs include. If you still need to hold onto the contents of the bucket but won’t be regularly accessing them, you have an excellent candidate on your hands for being archived into an appropriately tagged Glacier vault.
If a Usage Type is listed as “Requests-Tier1,” it represents PUT, COPY, POST, or LIST requests. If it’s listed as “Requests-Tier2,” it represents GET requests. If it’s listed as “Requests-Tier3,” it represents Glacier Archive and Restore requests. You can easily see which of your buckets are receiving which type of requests by keeping them tagged.
Using lifecycles to auto-archive
You can automate your process of archiving unaccessed S3s to Glacier using lifecycle rules for versioning on your S3 buckets. You can use a lifecycle rule to automatically move an item in S3 to Glacier after a certain amount of time.
It’s important to note that lifecycle rules apply to all of the data in an S3 bucket. This means that to employ auto-archival effectively, you’ll need to segregate your S3 buckets according to what type of data they contain and the according desired retention period. For example, you’ll likely want to store your log files in distinct S3s from other data that might be more regularly accessed—since log files are rarely accessed as time passes, the ability to auto-archive the S3 bucket in which they’re contained will help you stay efficient with your storage spending.
Ensuring efficient spending
From breaking down your storage costs by tag with an AWS Cost Allocation Report to visualizing the impact of your savings efforts over time with a Time Period Comparison, Cloudability can help you ensure deliberate and efficient spending on your AWS storage. Get started managing your storage costs and log in or sign up today.