}
	return target.isActive()
}

func (target *NSQTarget) isActive() (bool, error) {
	if target.producer == nil {
		producer, err := nsq.NewProducer(target.args.NSQDAddress.String(), target.config)
		if err != nil {
			return false, err
		}
		target.producer = producer
	}

	if err := target.producer.Ping(); err != nil {
		// To treat "connection refused" errors as errNotConnected.
		if xnet.IsConnRefusedErr(err) {

      assertEquals(0, Queues.drain(q, ImmutableList.of(), 1, 10, MILLISECONDS));

      Producer producer = new Producer(q, 1);
      // producing one, will ask for two
      Future<?> producerThread = threadPool.submit(producer);
      producer.beganProducing.await();

      // make sure we time out
      Stopwatch timer = Stopwatch.createStarted();

/target/
/work/
/bin/
/.settings/
.project
.classpath
*.iml
.idea
.DS_Store

      }
    },
    REDUCE_LAST {
      @Override
      Object operate(Stream<?> stream) {
        return stream.reduce((a, b) -> b);
      }
    },
    REDUCE_LAST_PARALLEL {
      @Override
      Object operate(Stream<?> stream) {
        return stream.parallel().reduce((a, b) -> b);
      }
    };

    abstract Object operate(Stream<?> stream);
  }

  @Param private Operation operation;

Doing this provides the following specific benefits:
- **Reduce the API surface** - We no longer need to maintain two methods.
- **Consistency** - All languages have consistent access to the same APIs and ergonomics in the DSL.
- **Reduce the size of the Gradle distribution** - We no longer need to carry multiple standard libraries for different languages.

## Decision

   * There is no need for a blocking lock; when the lock is not
   * available, threads try other slots (or the base).  During these
   * retries, there is increased contention and reduced locality,
   * which is still better than alternatives.
   *
   * Per-thread hash codes are initialized to random values.
   * Contention and/or table collisions are indicated by failed

This container branches off from
/tensorflow/tools/tf_sig_build_dockerfiles/. However, since
hermetic CUDA and hermetic Python is now available for Tensorflow, a lot of the
requirements installed on the original container can be removed to reduce the
footprint of the container and make it more reusable across different ML

	config.Producer.Retry.Max = 2
	config.Producer.Retry.Backoff = (1 * time.Second)
	config.Producer.Return.Successes = true
	config.Producer.Return.Errors = true
	config.Producer.RequiredAcks = 1
	config.Producer.Timeout = (5 * time.Second)
	// Set Producer Compression
	cc, ok := codecs[strings.ToLower(args.Producer.Compression)]
	if ok {
		config.Producer.Compression = cc

/target/
/.settings/
/fess-crawler*/target/
/fess-crawler*/bin/
/fess-crawler*/mydbflute/
/fess-crawler*/.project
/fess-crawler*/.classpath
/fess-crawler*/.settings/
/fess-crawler*/dependency-reduced-pom.xml
/fess-crawler*/dbflute_crawler/log/*.log
/fess-crawler*/dbflute_crawler/schema/project-schema-robot.xml
/fess-crawler/phantomjsdriver.log
.project
.idea

	// refer https://github.com/IBM/sarama/issues/765#issuecomment-254333355
	sconfig.Producer.Retry.Max = 2
	sconfig.Producer.Retry.Backoff = (10 * time.Second)
	sconfig.Producer.Return.Successes = true
	sconfig.Producer.Return.Errors = true
	sconfig.Producer.RequiredAcks = 1
	sconfig.Producer.Timeout = (10 * time.Second)
	sconfig.Net.ReadTimeout = (10 * time.Second)
	sconfig.Net.DialTimeout = (10 * time.Second)