Amazon EFS File System
Does ZopNight manage Amazon EFS File System?
Amazon EFS bills per GB-month by storage class, and the gap matters: the Standard class costs several times the Infrequent Access rate for data nobody reads. ZopNight tracks 90 days of EFS metrics, discovers replication configurations alongside each file system, and recommends lifecycle policies that move cold files to cheaper classes.
Rules that fire on Amazon EFS File System
At a glance
| Field | Value |
|---|---|
| Scheduling notes | discovery, metrics, cost tracking, and recommendations only. |
Amazon Elastic File System (EFS) provides shared NFS file storage billed per GB-month by storage class. File systems without lifecycle management keep rarely accessed data in the Standard class at several times the cost of Infrequent Access.
Storage class sets the rate
EFS bills for bytes stored, per GB-month, with no provisioning of capacity: the file system grows and shrinks with its contents. The rate depends on class. Standard is the default and the most expensive, while Infrequent Access and Archive hold the same data far cheaper but add a small per-GB charge on reads. Throughput has its own meter in provisioned and elastic modes; bursting throughput comes included with Standard storage.
Lifecycle is the whole game
Because the meter is pure GB-month, EFS cost optimization is placement. A lifecycle policy automatically moves files untouched for a chosen period, anywhere from 7 to 90 days, into Infrequent Access, and can move them back on access. File systems created without one, which is the common case since it is a setting nobody owns, hold years of cold files at the Standard rate for no benefit.
Patterns worth catching
File systems mounted by nothing, storing data for applications that no longer exist. The file system outlives its consumers because deleting shared storage feels risky. Cold datasets sitting in Standard because lifecycle was never enabled. And provisioned throughput configured for a one-time migration and never dialed back down.
Discovery, including replication
ZopNight discovers file systems on the 6-hour cycle, including cross-region replication configurations, which are found alongside the file system. Replication doubles storage cost by design, so knowing it exists matters for attribution. Hourly CloudWatch metrics from the EFS namespace run with a 90-day lookback, per-file-system cost comes from Cost Explorer or CUR 2.0, and recommendations cover lifecycle and tiering. There is no stop operation; savings come from class placement and deletion.
The class split in the console
EFS console, then File systems. Each file system’s detail page shows size split by storage class, Standard against IA, which is exactly the split a lifecycle policy exists to change.