Endpoints and regions
Zata has two regions, both in India. Each region has its own endpoint. Point your tool at the endpoint of the region where your bucket is.
Settings for any S3 tool
Always use https://. A request to http:// is redirected to https://, and many tools fail after the redirect.
One region per bucket, one region per key
- A bucket is in one region, the one you chose when you created it. You can’t move a bucket to the other region. To move data, copy it into a bucket in the other region, for example with rclone.
- An access key works in one region only. An Indore key gets
InvalidAccessKeyIdorAccessDeniedon the Mumbai endpoint, and the other way round. If you use both regions, create a key in each. See Access keys.
To see a bucket’s region, open Buckets in the console and look at the Region column. S3 tools don’t show it: aws s3 ls lists names and dates only.
To check from the command line, run head-bucket with a profile set up for one region. No output means the bucket is in that region. A 404 error means it isn’t there, or the name is wrong.
Path-style and virtual-host style
Both work. Use path-style when a bucket name contains a dot (.): in virtual-host style such names fail the certificate check.
Zata uses public Let’s Encrypt certificates for idr01.zata.ai, bom01.zata.ai and their subdomains. Every operating system trusts them, so you never need a custom certificate. Don’t turn off certificate checks in your tool.
Differences between the regions
Both regions offer the same features with one exception: in lifecycle rules, the options to keep only the newest versions and to filter by object size are available only in Mumbai.
Your plan covers both regions. Storage you use in Indore and Mumbai counts towards the same plan. See Billing.
If something doesn’t connect
See Wrong endpoint or region for NoSuchBucket, InvalidAccessKeyId and redirect errors.