Aias00 opened a new issue, #6574: URL: https://github.com/apache/shenyu/issues/6574
## Description `keepAlive` defaults to `false`. `HttpClient.keepAlive(false)` sends `Connection: close` on every outbound request, so connections are never reused. The comment says it was set to fix a Reactor Netty `Connection reset by peer` bug (reactor-netty#388, from 2019, long fixed in modern versions). ## Location ``` shenyu-plugin-httpclient/.../config/HttpClientProperties.java:117 (private boolean keepAlive; default false) HttpClientFactory.java:142 (httpClient.keepAlive(properties.isKeepAlive())) ``` ## Impact Every proxied request opens a new TCP connection (and new TLS handshake if HTTPS). At 10K RPS = 10K connections/sec churn, adding ~1ms TCP + ~50-100ms TLS latency per request, increasing kernel overhead (TIME_WAIT, conntrack). Makes the connection pool useless. ## Suggested fix Default `keepAlive = true`. If the RST issue resurfaces for specific upstreams, make it per-route config, not a global default. ## Related existing issue(s) None. Distinct from FD leaks — this is a default-config latency/throughput problem. _Identified during the 2026-08-02 audit; full list in [`docs/issue-candidates-2026-08-02.md`](docs/issue-candidates-2026-08-02.md)._ -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
