TATUM API - Recommended Connection Timeout Settings
When working the Tatum API, you may notice that our responses include the HTTP header:
Connection: keep-alive
This indicates that persistent HTTP connections are supported, allowing multiple requests to reuse the same underlying TCP connection for improved performance and reduced latency.
Tatum API - Request Timeout
- The request timeout on Tatumβs API is 60 seconds.
- This represents the maximum duration allowed for a request to be processed.
- If the request cannot be completed within that time window, it will be terminated automatically.
Attention
The connection timeout described below applies to Tatum API endpoints, not to Gateway JSON-RPC calls.
About the Keep-Alive
Header
Keep-Alive
Header- This header is optional and purely informational.
- Some clients may expect an additional header of the form:
Keep-Alive: timeout=N, max=M
- The
HTTP/1.1
standard does not require servers to include it, and its absence does not affect connection reuse or stability. - Tatumβs API supports keep-alive connections without explicitly specifying timeout or max values.
Client Configuration Recommendations
To maintain efficient and stable performance when connecting to Tatum APIs:
- Configure your HTTP client to reuse existing connections rather than opening new ones for every request.
- Set a reasonable idle connection timeout on your client, for example, 20 seconds, to prevent connections from remaining open indefinitely.
- Ensure your client manages its own connection pool and defines sensible limits for concurrent requests and idle connections.
Summary
Setting | Description | Recommendation |
---|---|---|
Request timeout | Max duration for a single request to complete | 60 seconds |
Keep-Alive header | Optional informational field (timeout , max ) | Not required |
Idle connection timeout (client) | How long to keep connections open when idle | ~20 seconds |
Connection reuse | Maintain persistent connections across requests | Enabled |
Note
Setting your clientβs connection pool and idle timeout parameters ensures efficient performance and prevents unnecessary reconnects when using Tatum APIs.
Updated about 5 hours ago