the index has 1 shards or 2+ shards, then we shouldn't just test against an
index with a random number of shards: there should be one test for the 1-shard
case, and another test for the 2+ shards case.

==== Abuse randomization in multi-threaded tests

Multi-threaded tests are often not reproducible due to the fact that there is
no guarantee on the order in which operations occur across threads. Adding

   * CacheBuilder#maximumSize} or {@link CacheBuilder#maximumWeight}.
   */
  SIZE {
    @Override
    boolean wasEvicted() {
      return true;
    }
  };

  /**
   * Returns {@code true} if there was an automatic removal due to eviction (the cause is neither
   * {@link #EXPLICIT} nor {@link #REPLACED}).
   */
  abstract boolean wasEvicted();

   *       that's not annotated nullable should throw {@link NullPointerException}.
   *   <li>If there is any non-private constructor or non-private static factory method declared by
   *       {@code cls}, all non-private instance methods will be checked too using the instance
   *       created by invoking the constructor or static factory method.
   *   <li>If there is any non-private constructor or non-private static factory method declared by

MinIO when the user is redirected to the OpenID Provider in the beginning. For some setups, extra configuration may be required for this step to work correctly.

For a simple setup where the user/client app accesses MinIO directly (i.e. with no intervening proxies/load-balancers), and each MinIO server (if there are more than one) has a unique domain name, this redirection should work automatically with no further configuration. For example, if the MinIO service is being accessed by the browser...

    for (int i = 0; i + 1 < asRanges.size(); i++) {
      Range<C> range1 = asRanges.get(i);
      Range<C> range2 = asRanges.get(i + 1);
      assertFalse(range1.isConnected(range2));
    }

    // test that there are no empty ranges
    for (Range<C> range : asRanges) {
      assertFalse(range.isEmpty());
    }

    // test that the RangeSet's span is the span of all the ranges

  }

  public void testNullPointers() {
    NullPointerTester tester = new NullPointerTester();
    tester.testAllPublicConstructors(Atomics.class); // there aren't any
    tester.testAllPublicStaticMethods(Atomics.class);
  }

sleep 1m

docker system prune -f || true
docker volume prune -f || true
docker volume rm $(docker volume ls -q -f dangling=true) || true

# Stop two nodes, one of each pool, to check that all S3 calls work while quorum is still there
[ "${MODE}" == "pools" ] && docker-compose -f minio-${MODE}.yaml stop minio2
[ "${MODE}" == "pools" ] && docker-compose -f minio-${MODE}.yaml stop minio6

# Pause one node, to check that all S3 calls work while one node goes wrong

- Don’t change public API lightly, avoid if possible, and include your reasoning in the PR if essential.  It causes pain for developers who use OkHttp and sometimes runtime errors.
- Favour a working external library if appropriate.  There are many examples of OkHttp libraries that can sit on top or hook in via existing APIs.
- Get working code on a personal branch with tests before you submit a PR.

   *
   * If a frame cannot be sent - because there are none enqueued or because the web socket is not
   * connected - this does nothing and returns false. Otherwise, this returns true and the caller
   * should immediately invoke this method again until it returns false.
   *
   * This method may only be invoked by the writer thread. There may be only thread invoking this
   * method at a time.
   */

import org.junit.jupiter.api.Test;

import static org.junit.jupiter.api.Assertions.assertEquals;

/**
 * A test which demonstrates maven's recursive inheritance where
 * a distinct value is taken from each parent contributing to
 * the final model of the project being assembled. There is no
 * overriding going on amongst the models being used in this test:
 * each model in the lineage is providing a value that is not present