How limits are applied
Most limits are applied on a per-server basis. Each Bunny Storage region runs multiple servers, and each server tracks limits independently. In practice, this allows you to achieve much higher concurrency than the per-server limits suggest.Connection limits
Request limits by operation
Certain operations have their own concurrency and size limits:HTTP API limits
The HTTP API is limited to 250 concurrent connections per zone, per server. Connections beyond this limit are rejected until capacity frees up.S3 API throughput and rate limits
To move data faster than a single connection allows, split transfers across multiple parallel connections rather than relying on one.
Rate limit responses
The response you receive depends on which limit you reach and which protocol or API you’re using.
For the per-IP FTP response, N is the number of connections currently open from your IP.
If you hit one of these limits:
- Reduce concurrent connections
- Implement exponential backoff retry logic (and honor the
Retry-Afterheader when present) - Distribute requests across longer time periods
File and directory naming
When naming files and folders in Bunny Storage:- A file name including its full path cannot exceed 6,000 characters.
- A directory path cannot exceed 6,000 characters.
- File names cannot start or end with a space, and cannot be empty.
SFTP path restrictions
For security, SFTP paths cannot contain.. (parent-directory traversal). If a path includes .., the SFTP connection fails silently, with no error message. If your SFTP client stops working unexpectedly, check that none of your paths contain ...