-
Fixed an issue where the write timeout incorrectly used the read timeout when
max_idle_timewas set. -
Fixed an invalid warning caused by the exit of old processes during asynchronous reload.
-
Refactored server process management logic.
-
Optimized handling of fork failures: added a retry mechanism with a 1-second wait before each retry to reduce excessive error logging.
-
Changed the
closedfield in the underlyingConnectionstruct to use atomic operations, resolving a potential data race condition. -
Fixed an integer overflow issue in the Redis server.
-
修复当设置
max_idle_time时,写超时错误地使用了读超时时间的问题。 -
修复异步重载期间因旧进程退出而导致的无效警告。
-
重构服务器进程管理逻辑。
-
优化进程 fork 失败后的处理:增加重试机制,并在重试前等待 1 秒,避免产生过多错误日志。
-
将底层
Connection结构体的closed字段改为原子操作,解决潜在的数据竞争问题。 -
修复 Redis 服务端中的整型溢出问题。