Skip to main content

Object Storage overview

S3-compatible buckets managed from the dashboard, accessed with standard S3 clients and SDKs.

SailOps Object Storage gives each project buckets backed by a managed S3-compatible endpoint. Upload, download and manage objects with any S3-compatible tool — the AWS CLI, Boto3, the AWS SDK, MinIO client, or any library that speaks S3.

Key features​

  • S3-compatible API — Path-style addressing, multipart uploads, versioning, lifecycle rules
  • Visibility control — Buckets are private by default; optionally enable public read so anyone with an object URL can download it
  • Per-bucket access rules — Fine-grained S3 action grants bound to service users
  • Service users and keys — Create dedicated identities with scoped access keys (up to two active keys per user)
  • Object browser — List, upload, download and delete objects from the dashboard
  • Versioning — Enable or suspend versioning per bucket; once enabled, it can be suspended but not disabled
  • Lifecycle rules — Set retention for current objects, previous versions, and incomplete multipart uploads
  • Usage tracking — Per-bucket metered billing for storage, reads, writes and downloads
  • Resource tags — Organize buckets with custom key-value tags

Availability​

Object Storage requires an activated billing account. It is unavailable during the free trial.

Create a bucket​

  1. Open Object Storage in the sidebar
  2. Click Create bucket
  3. Enter a bucket name (3–63 lowercase letters, digits or hyphens; permanent and globally unique on the endpoint)
  4. Click Create bucket

The bucket starts private with versioning disabled. Incomplete multipart uploads are cleaned up after seven days by default; completed objects do not expire.

Connection details​

Each bucket exposes standard S3 connection details:

PropertyValue
S3 endpointShown on the bucket overview page
Signing regionShown on the bucket overview page
AddressingPath-style (forcePathStyle: true in your SDK)
AuthenticationAccess key and secret key from a service user

Quick start with the AWS CLI​

aws --endpoint-url <endpoint> --region <region> s3 ls s3://<bucket-name>/

Replace <endpoint>, <region> and <bucket-name> with the values shown on the bucket overview.

Service users and keys​

Service users are S3 identities scoped to your project. Each user can hold up to two active keys, allowing zero-downtime rotation.

  1. Open Object Storage → Users & keys
  2. Click Create user and enter a name
  3. Once the user is provisioned, click Create key
  4. Save the access key and secret key — the secret is shown only once

To rotate credentials: create a replacement key, update your applications, then revoke the old key.

Users can be suspended (all their keys stop working) or deleted (all keys are permanently revoked).

Each user card also shows a Bucket access summary listing every bucket the user can reach, the scope of each rule (whole bucket or a prefix) and the permission level. Users with no access rules display a warning — their keys cannot reach any bucket until a rule is added from the bucket's Access tab.

Access rules​

Access is controlled per bucket on the Access tab. Each rule binds a service user to a set of S3 actions on the whole bucket or a prefix:

ActionDescription
s3:GetObjectRead objects
s3:GetObjectVersionRead previous versions
s3:PutObjectWrite objects
s3:DeleteObjectDelete the current version
s3:DeleteObjectVersionPermanently erase a specific version
s3:AbortMultipartUploadAbort an incomplete multipart upload
s3:ListMultipartUploadPartsList parts of an incomplete multipart upload

Each rule also controls whether the user can list objects and list versions within the scope. Listing multipart uploads is only available for whole-bucket grants — prefix-scoped rules cannot include this permission. Every user with a rule on the bucket can check that it exists (S3 HeadBucket), as tools such as rclone do at startup. This does not let a prefix-scoped user list objects outside its prefix, but with the read-stats parameter it reports the bucket's total object count and size.

Bucket settings​

Visibility​

ModeBehaviour
PrivateAccess requires an authorized S3 key or a signed URL. This is the default for new buckets.
Public readAnyone with an object URL can read it. Listing, previous versions and writes remain protected. Reads and downloads count toward this bucket's usage.

To enable public read:

  1. Open the bucket's Settings tab
  2. Under Visibility, select Public read
  3. Click Save changes
  4. Confirm in the dialog that current objects will be publicly readable

You can switch back to private at any time without confirmation. A warning banner appears while changes are still being applied.

When a bucket is public-read and ready, a Copy public URL button appears next to each object in the browser.

warning

Enabling public read makes every current object in the bucket downloadable by anyone who knows its URL. Downloads are billed to the bucket owner.

Versioning​

StateMeaning
DisabledOnly the initial state of a new bucket
EnabledEvery put creates a new version; deletions create a delete marker
SuspendedNew puts overwrite the current version; existing versions are preserved

Once enabled, versioning cannot be disabled — only suspended.

Lifecycle​

  • Incomplete multipart uploads — Abort after a configurable number of days (1–365)
  • Current object retention — Expire current objects after a number of days (0 = keep indefinitely)
  • Previous version retention — Permanently delete previous versions after a number of days (0 = keep indefinitely)

Object browser​

The Objects tab lets you browse, upload and download objects from the dashboard:

  • Search by prefix and navigate folders
  • Toggle Include versions to see all versions and delete markers
  • Upload a file directly from the browser
  • Download or delete individual objects and versions

Connections​

You can connect a bucket to containers and jobs so SailOps automatically injects S3 credentials and settings into the workload. Each connection creates a dedicated S3 identity scoped to the bucket with its own access key pair.

Creating a connection from the bucket page​

  1. Open the bucket and go to the Connections tab
  2. Click Create connection
  3. Select the workload (container or scheduled job) to connect
  4. Enter an Alias (default AWS)
  5. Choose the Access level (readonly, readwrite, or full)
  6. Click Create Connection

You can also create Object Storage connections from the container or job Connections tab by selecting Object Storage as the target type.

Injected environment variables​

Variable (prefixed with alias)Description
ACCESS_KEY_IDS3 access key for this connection
SECRET_ACCESS_KEYS3 secret key for this connection
ENDPOINT_URLS3 endpoint URL
ENDPOINT_URL_S3S3 endpoint URL (SDK-specific)
REGIONS3 signing region
DEFAULT_REGIONS3 signing region (SDK-specific)
BUCKET_NAMETarget bucket name
S3_FORCE_PATH_STYLEAlways true
tip

Use AWS as the alias so standard AWS SDKs pick up the credentials automatically — AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, AWS_ENDPOINT_URL_S3, AWS_DEFAULT_REGION.

Access levels​

LevelPermissions
readonlyRead objects, previous versions and listings
readwriteRead, upload and delete objects
fullRead, write and permanently delete individual versions
warning

The full level grants permanent version deletion. Bucket settings, policies and user administration are never granted through connections.

Lifecycle​

  • Containers restart when connection variables change; scheduled jobs use the new values on their next run.
  • Deleting a connection revokes its S3 credentials and removes the injected variables. The bucket and its objects are kept.
  • Deleting a bucket automatically disconnects all connected workloads.

Deleting a bucket​

Two options are available on the Settings tab:

  • Delete empty bucket — Removes the bucket only if it contains no objects, versions or incomplete uploads
  • Purge bucket — Permanently deletes the bucket and all its contents, including previous versions. Requires typing the bucket name to confirm.

Both confirmation dialogs show an impact summary listing the service users that will lose access and the connected workloads whose credentials will be revoked. Review this list before confirming.

Billing​

Object Storage uses metered billing with four dimensions:

DimensionUnitDescription
StorageGiB-monthTime-weighted average volume, including versions
Writes & listingsRequestsPUT, POST, DELETE and list operations
Reads & HEADRequestsGET and HEAD operations
DownloadsGiBEgress bytes (same rate from SailOps internal network and the Internet)

View per-bucket usage and costs on the bucket's Usage & costs tab, or see the project-wide breakdown from Object Storage → Usage & costs.