simultaneous connections wastes resources; too few harms latency.

For asynchronous calls, [`Dispatcher`](https://square.github.io/okhttp/4.x/okhttp/okhttp3/-dispatcher/) implements policy for maximum simultaneous requests. You can set maximums per-webserver (default is 5), and overall (default is 64)....

awscli s3api put-object-lock-configuration --bucket mybucket --object-lock-configuration 'ObjectLockEnabled=\"Enabled\",Rule={DefaultRetention={Mode=\"GOVERNANCE\",Days=1}}'
```

### Set object lock

PutObject API allows setting per object retention mode and retention duration using `x-amz-object-lock-mode` and `x-amz-object-lock-retain-until-date` headers. This takes precedence over any bucket object lock configuration w.r.t retention.

```sh

    assertEvents(
        "R0.00", // First comes the saved-up burst, which defaults to a 1-second burst (2 requests).
        "R0.00", "R0.00", // Now comes the free request.
        "R0.50", // Now it's 0.5 seconds per request.
        "R0.50");

    limiter.setRate(Double.POSITIVE_INFINITY);
    limiter.acquire();
    limiter.acquire();
    limiter.acquire();

    assertEvents(
        "R0.00", // First comes the saved-up burst, which defaults to a 1-second burst (2 requests).
        "R0.00", "R0.00", // Now comes the free request.
        "R0.50", // Now it's 0.5 seconds per request.
        "R0.50");

    limiter.setRate(Double.POSITIVE_INFINITY);
    limiter.acquire();
    limiter.acquire();
    limiter.acquire();

 * This is for the benefit of container environments that implement code unloading.
 *
 * Most applications should share a process-wide [TaskRunner] and use queues for per-client work.
 */
class TaskRunner(
  val backend: Backend,
  internal val logger: Logger = TaskRunner.logger,
) : Lockable {
  private var nextQueueName = 10000
  private var coordinatorWaiting = false

//   # is local and target is remote which is also minio.
//   snowball:
//     disable: false # optionally turn-off snowball archive transfer
//     batch: 100 # upto this many objects per archive
//     inmemory: true # indicates if the archive must be staged locally or in-memory
//     compress: true # S2/Snappy compressed archive
//     smallerThan: 5MiB # create archive for all objects smaller than 5MiB

	XMLName xml.Name `xml:"ReplicationConfiguration" json:"-"`
	Rules   []Rule   `xml:"Rule" json:"Rules"`
	// RoleArn is being reused for MinIO replication ARN
	RoleArn string `xml:"Role" json:"Role"`
}

// Maximum 2MiB size per replication config.
const maxReplicationConfigSize = 2 << 20

// ParseConfig parses ReplicationConfiguration from xml
func ParseConfig(reader io.Reader) (*Config, error) {
	config := Config{}

 * type is determined by the crawler data serializer configuration.
 * </p>
 * <p>
 * The class is thread-safe and uses ThreadLocal to maintain Kryo instances
 * per thread to avoid synchronization overhead.
 * </p>
 *
 */
public class DataSerializer {

    /** Logger for this class. */
    private static final Logger logger = LogManager.getLogger(DataSerializer.class);

    /**
     * Sets the URL stream handler factory for the environment. This
     * allows specification of the factory used in creating underlying
     * stream handlers. This can be called once per JVM instance.
     *
     * @param factory
     *            The URL stream handler factory.
     */
    public static void setURLStreamHandlerFactory(final URLStreamHandlerFactory factory) {

        }

        @Test
        void rootUrlIsRoot() throws Exception {
            DummySmbResourceLocator root = new DummySmbResourceLocator("smb://server/");
            // last component includes trailing slash per specification
            assertEquals("server/", root.getName());
            assertEquals("smb://", root.getParent());
            assertTrue(root.isRoot());
        }

        @Test