Migrate from Google Cloud Storage to Tigris | Tigris Object Storage Documentation

On this page

You can migrate from Google Cloud Storage (GCS) to Tigris with zero downtime using Tigris's shadow bucket feature. Tigris is S3-compatible, so applications using GCS's S3-compatible XML API need only an endpoint and credential change.

Why migrate from GCS to Tigris?

Migration approach

Tigris supports lazy migration using shadow buckets. Instead of copying all your data upfront, Tigris fetches objects from your GCS bucket on demand and caches them for future access. No downtime required.

You can also enable write-through mode, which syncs new writes back to your GCS bucket. This means your existing GCS bucket stays up to date throughout the migration, and you can take as long as you need before completing the cutover.

Prerequisites

Before starting, make sure you have:

Step 1: Create GCS HMAC credentials

Tigris uses S3-compatible APIs, so you need HMAC credentials (not OAuth tokens) for GCS access.

  1. Go to the Google Cloud Console Interoperability page.
  2. If you don't already have a service account with access to your GCS bucket, create one in the IAM & Admin section. Assign it Storage Object Viewer for read-only migration, or Storage Object Admin if you plan to use write-through mode.
  3. On the Interoperability page, find the service account and click Create Key under the HMAC Keys section.
  4. Copy and securely store the Access Key and Secret Key.

Step 2: Configure the shadow bucket

  1. Go to the Tigris Dashboard
  2. Click Buckets in the left menu
  3. Select the bucket you want to migrate data into
  4. Click Settings
  5. Find Enable Data Migration and toggle it on
  6. Enter your GCS connection details:
    • Endpoint: https://storage.googleapis.com
    • Region: auto (GCS does not require a region for S3-compatible access)
    • Access Key ID: Your GCS HMAC access key from Step 1
    • Secret Access Key: Your GCS HMAC secret key from Step 1
    • Bucket: The name of your GCS source bucket

The CLI flow is two commands: configure the shadow source, then optionally drain it.

1. Configure the shadow bucket with tigris buckets set-migration:

tigris buckets set-migration my-bucket \

--bucket source-bucket \

--endpoint https://storage.googleapis.com \

--region auto \

--access-key <hmac-access-key> \

--secret-key <hmac-secret-key>

Add --write-through for write-through mode, or --disable to clear the migration configuration.

2. Actively migrate (optional). Lazy migration only copies objects when they're requested. To migrate every remaining object server-side, run tigris buckets migrate:

tigris buckets migrate my-bucket

The command runs in the foreground and reports progress as it goes.

Step 3: Update your application

Point your application to Tigris by updating the endpoint and credentials in your S3 client configuration.

Using the AWS CLI:

aws s3 ls s3://your-bucket/ \

--endpoint-url https://t3.storage.dev \

--region auto

Or with boto3:

import boto3

s3 = boto3.client(

"s3",

endpoint_url="https://t3.storage.dev",

region_name="auto",

aws_access_key_id="your-tigris-access-key",

aws_secret_access_key="your-tigris-secret-key",

)

The only changes needed are the endpoint URL, region, and credentials. Your existing bucket names, object keys, and API calls stay the same.

Step 4: Verify the migration

Once your application points to Tigris, objects are migrated on first access. To verify:

  1. Request an object that exists in your GCS bucket
  2. Confirm it returns successfully through Tigris
  3. Check that subsequent requests are served directly from Tigris

Step 5: Enable write-through (optional)

To keep your GCS bucket in sync during the migration, enable write-through in the shadow bucket settings. With write-through enabled:

This keeps your GCS bucket current so you can fall back at any point.

Step 6: Complete the migration

Once your workloads are running well on Tigris, disable the shadow bucket configuration. Tigris becomes your primary object store.

FAQ

Does migration require downtime?

No. Shadow bucket migration happens transparently. Your application continues serving requests while objects are migrated on first access.

Do I need to change my application code?

Only the endpoint URL, region, and credentials. Tigris supports the S3 API, so your existing bucket names, object keys, and API calls work without changes.

Can I roll back to GCS?

Yes. If you enable write-through mode, your GCS bucket stays in sync with all new writes. You can switch back to GCS at any point by reverting your endpoint configuration.

Why do I need HMAC credentials instead of OAuth?

Tigris uses S3-compatible APIs for migration. GCS supports S3-compatible access through its Interoperability API, which requires HMAC-style credentials rather than the OAuth tokens used by GCS's native JSON API.

What happens to objects I haven't accessed yet?

They remain in your GCS bucket. Tigris only copies objects when they're first requested. Objects that are never accessed are never transferred.

How much does Tigris cost compared to GCS?

Tigris charges $0.02/GB/month for standard storage with zero egress fees. See the Tigris pricing page for full details.