在rails中捕获超时异常似乎跳过了救援并且"一直向上"爆炸

Question

我有一个控制器:

class EventsController < ApplicationController  
  def index
    begin
      SystemTimer.timeout_after(10, CustomTimeoutError) do
        sleep(11)
      end
    rescue CustomTimeoutError => e
      # swallow
    end
  end
end

出于某种原因,救援语句不要再追超时,而是它的冒泡一路攀升到顶部,堆栈跟踪被倾倒到控制台等使用defatult超时::错误有同样的效果.这只发生在生产上,而不是在我的开发机器上.就好像其他东西正在观察超时并在事件到达我的救援之前捕获它们.

产生的堆栈跟踪是这样的:

[GEM_ROOT]/gems/SystemTimer-1.2/lib/system_timer/concurrent_timer_pool.rb:63:in `read_reply'

vendor/gems/redis-1.0.4/lib/redis/client.rb:444:in `process_command'

vendor/gems/redis-1.0.4/lib/redis/client.rb:442:in `map'

vendor/gems/redis-1.0.4/lib/redis/client.rb:442:in `process_command'

vendor/gems/redis-1.0.4/lib/redis/client.rb:431:in `raw_call_command'

vendor/gems/redis-1.0.4/lib/redis/client.rb:452:in `call'

vendor/gems/redis-1.0.4/lib/redis/client.rb:452:in `maybe_lock'

vendor/gems/redis-1.0.4/lib/redis/client.rb:428:in `raw_call_command'

vendor/gems/redis-1.0.4/lib/redis/client.rb:332:in `call_command'

vendor/gems/redis-1.0.4/lib/redis/client.rb:381:in `method_missing'

vendor/gems/ohm-0.0.35/lib/ohm/collection.rb:179:in `size'

vendor/gems/ohm-0.0.35/lib/ohm/collection.rb:65:in `empty?'

vendor/gems/ohm-0.0.35/lib/ohm/collection.rb:33:in `sort'

vendor/gems/ohm-0.0.35/lib/ohm/collection.rb:48:in `first'

vendor/gems/ohm-0.0.35/lib/ohm.rb:129:in `first'

lib/twitter_helper.rb:58:in `get_twitter_searches'

lib/twitter_helper.rb:57:in `each'

lib/twitter_helper.rb:57:in `get_twitter_searches'

lib/twitter_helper.rb:100:in `get_twitter_searches_or_messages'

app/controllers/events_controller.rb:66:in `show'

[GEM_ROOT]/gems/SystemTimer-1.2/lib/system_timer.rb:56:in `timeout_after'

app/controllers/events_controller.rb:65:in `show'

vendor/rails/actionpack/lib/action_controller/base.rb:1331:in `send'

Answer 1

您是如何声明自定义 CustomTimeoutError 类的？它是 Exception 还是 StandardError 的直接后代？当从 Exception 下降时，我确切地看到了这种行为，并养成了始终使我的错误类从 StandardError 下降的习惯。