Alex Zhang

@fankeke 是的。 目前 `resty-sync` 的做法能够完全避免掉这些问题。

@fankeke 你可以仔细看下 `_M.start` 这个函数，我通过在共享内存里设置一个特殊的标记，即用 `LOCK_TIMER_KEY` 作为 key，存放成功启动 sync timer 的 worker 的 id，过期时间为这个 `sync` 实例所设置的 `interval ` + 10，且每次定时器在处理回调的时候，都会更新下这个 key ，以避免过期。 如果某个 worker 进入到这个函数，先第一次抢锁，抢到锁的，首先会判断这个 `LOCK_TIMER_KEY` 对应的值是不是存在，如果不存在，则由这个 worker 抢占到 timer，同时共享内存里的 `LOCK_TIMER_KEY`...

@fankeke 恩，这个 `new` 操作你要放到 init 阶段，可以放在 init 阶段再试试

See the relevant issue: https://github.com/openresty/stream-lua-nginx-module/issues/167 for details.

> So ngx_http_lua_module also needs a patch. OK, I will push a patch to that repo.

> Look like the original `ngx_http_lua_check_broken_connection` is modified from `ngx_http_upstream_check_broken_connection` in 2012. In 2013, Nginx added the `NGX_USE_EPOLL_EVENT` check to fix https://trac.nginx.org/nginx/ticket/320. > However, this code doesn't synchronize the fix....

> @tokers > `ngx_http_lua_check_broken_connection` doesn't have this patch because it is written in 2012, > while this patch is added to `ngx_http_lua_check_broken_connection` in 2013. OK 👌

I have written a patch to solve this problem, I will push it to the meta-lua-nginx-module repo.

> I have written a patch to solve this problem, I will push it to the meta-lua-nginx-module repo. Nonetheless, I don't know why the `ngx_stram_lua_check_broken_connection ` would be triggerd duplicately...

@dndx Any idea about this?