AWS makes it cheaper to store little-used data with EFS Archive

Amazon’s AWS has introduced a new storage class that makes it cheaper to store data that is rarely touched.

Dubbed EFS Archive, it becomes the third storage class behind EFS Standard (where data is stored on SSDs to minimise latency for actively used data) and EFS Infrequent Access (which is a cheaper tier for data that’s accessed only a few times every quarter).

EFS Archive is intended for “cold”, long-lived data that is accessed only a few times a year. AWS suggests it might be used for data such as medical records, data that’s retained for regulatory compliance or machine learning datasets. The latter of those options is particularly pertinent, as Amazon has spent much of this week’s AWS re:Invent conference talking about Amazon Q, its new enterprise-grade AI assistant.

Although AWS claims EFS Archive offers similar performance to EFS Infrequent Access (EFS IA), it costs 50% less to store. The kicker is that it has a three times higher request charge.

EFS Archive policies

IT admins won’t have to manually move data between the different storage classes if they don’t want to. AWS lifecycle policies can be implemented so that data that hasn’t been accessed for 30 days is automatically tranistioned to EFS IA, while files that haven’t been touched for 90 days are plunged into EFS Archive.

No matter what class your files are in, they are all still instantly accessible from the same shared file system, so there’s no need to go hunting in different storage pools for your files. AWS claims that all files, no matter what the storage class, will have GB/sec throughput and eleven-nines durability.

EFS Archive will be available in all regions, except China; Amazon fails to offer an explanation as to why China should be excluded from the new storage class.

The full pricing for the different EFS tiers in each region can be found on the Amazon EFS Pricing site.