logstash-filter-grok 4.0.0 → 4.0.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: d06febb2d4c48dd7c02d2d89ecbe2c379c345b980687d10e80542192f3b232d9
-  data.tar.gz: ddb45461cff71c3fee2c7cd9e84780e0c0c216b207c23cee42ed39d499be25bf
+  metadata.gz: 4a821dcb7736046a21e53c4cbb4d23e9e807175c0641e4f6855674bca39e00cb
+  data.tar.gz: e1b8780ab43c0701dc8a4c6db7335da834fd6fb057957dbca8d61bb190c8585b
 SHA512:
-  metadata.gz: 7db250b93474764efb1c769aab63fe9e8850fa9eac4e14ad4b5ec82bdb296d6fe6edc92e4604c7f70cc52199929b2ce31b523cd07f75abb7fd781da64f74b385
-  data.tar.gz: 55051f431618fa87a944984de8b62ec81daaed169e719fad74d84d323b034d776d76bf26d568f5b587cabd16008ff2154c6b245f640287df36cf26ed7b7a0f9c
+  metadata.gz: ce2e26791c303cd15c504e136252cf01ced839481a2afd01536a76253b12f87f0d48826b522315f6b19dd0cdeefbb23ea9d7718e5dedfb3096539e077fd8f1d0
+  data.tar.gz: 386308027f66ec688f55712d83acca7f692f6b9fad3c0cbe0a4bc865e1beaa2c5d4c32757a572f388a73bede0788e11740ce989b84516da26e651095ed70cef1

data/CHANGELOG.md

@@ -1,3 +1,6 @@
+## 4.0.1
+  - Fix a potential race
+
 ## 4.0.0
   - Major performance improvements due to reduced locking
 

data/lib/logstash/filters/grok/timeout_enforcer.rb

@@ -1,44 +1,37 @@
 class LogStash::Filters::Grok::TimeoutEnforcer
-  attr_reader :running
-
   def initialize(logger, timeout_nanos)
     @logger = logger
-    @running = false
+    @running = java.util.concurrent.atomic.AtomicBoolean.new(false)
     @timeout_nanos = timeout_nanos
 
     # Stores running matches with their start time, this is used to cancel long running matches
     # Is a map of Thread => start_time
     @threads_to_start_time = java.util.concurrent.ConcurrentHashMap.new
-    @cancel_mutex = Mutex.new
+  end
+
+  def running
+    @running.get()
   end
 
   def grok_till_timeout(grok, field, value)
     begin
       thread = java.lang.Thread.currentThread()
-      start_thread_groking(thread)
+      @threads_to_start_time.put(thread, java.lang.System.nanoTime)
       grok.execute(value)
-    rescue InterruptedRegexpError => e
+    rescue InterruptedRegexpError, java.lang.InterruptedException => e
       raise ::LogStash::Filters::Grok::TimeoutException.new(grok, field, value)
     ensure
-      unless stop_thread_groking(thread)
-        @cancel_mutex.lock
-        begin
-          # Clear any interrupts from any previous invocations that were not caught by Joni
-          # It may appear that this should go in #stop_thread_groking but that would actually
-          # break functionality! If this were moved there we would clear the interrupt
-          # immediately after setting it in #cancel_timed_out, hence this MUST be here
-          java.lang.Thread.interrupted
-        ensure
-          @cancel_mutex.unlock
-        end
-      end
+      # If the regexp finished, but interrupt was called after, we'll want to
+      # clear the interrupted status anyway
+      @threads_to_start_time.remove(thread)
+      thread.interrupted
     end
   end
 
   def start!
-    @running = true
+    @running.set(true)
     @timer_thread = Thread.new do
-      while @running
+      while @running.get()
         begin
           cancel_timed_out!
         rescue Exception => e
@@ -54,49 +47,23 @@ class LogStash::Filters::Grok::TimeoutEnforcer
   end
 
   def stop!
-    @running = false
+    @running.set(false)
     # Check for the thread mostly for a fast start/shutdown scenario
     @timer_thread.join if @timer_thread
   end
 
   private
 
-  # These methods are private in large part because if they aren't called
-  # in specific sequence and used together in specific ways the interrupt
-  # behavior will be incorrect. Do NOT use or modify these methods unless
-  # you know what you are doing!
-
-  def start_thread_groking(thread)
-    # Clear any interrupts from any previous invocations that were not caught by Joni
-    java.lang.Thread.interrupted
-    @threads_to_start_time.put(thread, java.lang.System.nanoTime)
-  end
-
-  # Returns falsy in case there was no Grok execution in progress for the thread
-  def stop_thread_groking(thread)
-    @threads_to_start_time.remove(thread)
-  end
-
   def cancel_timed_out!
     now = java.lang.System.nanoTime # save ourselves some nanotime calls
-    @threads_to_start_time.entry_set.each do |entry|
-      start_time = entry.get_value
-      if start_time < now && now - start_time > @timeout_nanos
-        thread  = entry.get_key
-        # Ensure that we never attempt to cancel this thread unless a Grok execution is in progress
-        # Theoretically there is a race condition here in case the entry's grok action changed
-        # between evaluating the above condition on the start_time and calling stop_thread_groking
-        # Practically this is impossible, since it would require a whole loop of writing to an
-        # output, pulling new input events and starting a new Grok execution in worker thread
-        # in between the above `if start_time < now && now - start_time > @timeout_nanos` and
-        # the call to `stop_thread_groking`.
-        if stop_thread_groking(thread)
-          @cancel_mutex.lock
-          begin
-            thread.interrupt()
-          ensure
-            @cancel_mutex.unlock
-          end
+    @threads_to_start_time.forEach do |thread, start_time|
+      # Use compute to lock this value
+      @threads_to_start_time.computeIfPresent(thread) do |thread, start_time|
+        if start_time < now && now - start_time > @timeout_nanos
+          thread.interrupt
+          nil # Delete the key
+        else
+          start_time # preserve the key
         end
       end
     end

data/logstash-filter-grok.gemspec

@@ -1,7 +1,7 @@
 Gem::Specification.new do |s|
 
   s.name            = 'logstash-filter-grok'
-  s.version         = '4.0.0'
+  s.version         = '4.0.1'
   s.licenses        = ['Apache License (2.0)']
   s.summary         = "Parses unstructured event data into fields"
   s.description     = "This gem is a Logstash plugin required to be installed on top of the Logstash core pipeline using $LS_HOME/bin/logstash-plugin install gemname. This gem is not a stand-alone program"

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: logstash-filter-grok
 version: !ruby/object:Gem::Version
-  version: 4.0.0
+  version: 4.0.1
 platform: ruby
 authors:
 - Elastic
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2017-11-27 00:00:00.000000000 Z
+date: 2017-12-18 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   requirement: !ruby/object:Gem::Requirement