BIND9 递归查询超时机制

最新推荐文章于 2024-08-27 15:38:47 发布

最新推荐文章于 2024-08-27 15:38:47 发布 · 806 阅读

文章标签：

#运维

本文详细阐述了如何根据网络请求的失败次数调整重试间隔时间，包括初始重试时间设置、指数级增长的重试策略，以及确保重试等待时间不超过固定上限的逻辑。此实现旨在优化网络请求的可靠性与效率。

static inline void fctx_setretryinterval(fetchctx_t *fctx, unsigned int rtt) { unsigned int seconds; unsigned int us; /* * We retry every .8 seconds the first two times through the address * list, and then we do exponential back-off. */ if (fctx->restarts < 3) us = 800000; else us = (800000 << (fctx->restarts - 2)); /* * Double the round-trip time. */ rtt *= 2; /* * Always wait for at least the doubled round-trip time. */ if (us < rtt) us = rtt; /* * But don't ever wait for more than 10 seconds. */ if (us > 10000000) us = 10000000; seconds = us / 1000000; us -= seconds * 1000000; isc_interval_set(&fctx->interval, seconds, us * 1000); }