Cloud Storage Bucket

Cloud Storage buckets hold objects billed by storage class, volume, operations, and egress. Buckets without lifecycle rules quietly accumulate stale data in expensive storage classes for years.

The storage class sets the GB-month rate

Every object in a bucket carries a storage class: Standard, Nearline, Coldline, or Archive. The class determines the per-GB-month rate along with a minimum storage duration and a retrieval charge for the colder tiers. On top of stored bytes, the meter counts Class A and Class B operations (writes and list calls cost more than reads) and network egress when data leaves Google Cloud. Nothing about a bucket bills for existing; the money is entirely in what it holds and how often that data is touched. That shape makes the cost invisible day to day and cumulative over quarters.

Bucket metrics from Cloud Monitoring

ZopDev discovers buckets via Cloud Asset Inventory, then pulls each bucket’s total size and object count from Cloud Monitoring time series, so the platform knows not just that a bucket exists but how many bytes sit in it. Storage and operation spend is attributed from the billing export, and recommendation rules propose lifecycle policies and storage-class fixes. A bucket cannot be stopped, so there is no schedule to apply. The leverage is tiering: storage-class moves are the biggest recurring GCS saving.

Where object spend hides

Three patterns dominate. Standard-class data nobody has read in months, which Nearline or Coldline would hold for a fraction of the rate. Buckets with no lifecycle rules, where old versions, temp exports, and log dumps pile up forever because deletion requires a human to remember. And noncurrent object versions retained by versioning, each superseded version billing at the full class rate until a lifecycle rule expires it.

Walking the bucket list in the console

Google Cloud console → Cloud Storage → Buckets lists every bucket with its location and default storage class. Opening a bucket’s Configuration tab shows its lifecycle rules. An empty rules list on a bucket holding years of data is the fastest tell that tiering money is being left on the table.