Amazon S3 Bucket
Does ZopNight manage Amazon S3 Bucket?
An S3 bucket bills per GB-month by storage class, per 1,000 requests by request type, and per GB transferred out. Nothing stops or starts, so savings come from placement: ZopNight tracks 90 days of storage metrics per bucket and flags missing lifecycle rules, noncurrent-version pileups, and stale multipart uploads.
Rules that fire on Amazon S3 Bucket
At a glance
| Field | Value |
|---|---|
| Scheduling notes | discovery, metrics, cost tracking, and recommendations only. |
Amazon Simple Storage Service (S3) provides object storage billed per GB-month by storage class, plus request and transfer charges. Buckets without lifecycle policies keep cold data in the most expensive tier indefinitely.
Three charges on every bucket
Storage bills per GB-month at a rate set by storage class: Standard at the top, with Intelligent-Tiering, Infrequent Access, and the Glacier classes progressively cheaper. Requests bill per 1,000 by type, with PUT and LIST costing roughly 10 times GET. Data transfer out to the internet bills per GB. There is nothing to stop or schedule here; S3 cost is entirely a placement and hygiene problem.
What versioning does to the meter
With versioning enabled, overwriting or deleting an object does not free storage. Every noncurrent version keeps billing at the bucket’s class rate until a lifecycle rule expires it. A bucket with heavy overwrite traffic and no noncurrent-version expiry grows without bound while appearing to hold a constant object count.
The quiet accumulators
Three patterns produce most S3 waste: cold data sitting in Standard because no lifecycle policy was ever written; noncurrent versions piling up under versioning; and incomplete multipart uploads, the failed large uploads whose parts remain billable and invisible until an abort rule cleans them out. None of the three shows up in a console object count.
ZopNight’s view of a bucket
Buckets are discovered every 6 hours via Resource Explorer 2. S3’s CloudWatch storage metrics (bucket size and object count per storage class) are kept with a 90-day lookback, and per-bucket cost comes from Cost Explorer or CUR 2.0. Recommendations target storage-class tiering, missing lifecycle rules, Intelligent-Tiering candidates, and hygiene findings like multipart-upload accumulation.
Where buckets live in the console
S3 console, then Buckets, one global list rather than a per-region one, because bucket names are globally unique. The per-class storage breakdown for a bucket sits under Metrics on the bucket’s detail page, and it is the first place to look when a bucket’s bill and its apparent contents disagree.