bulkFuture.addListener(
        new Runnable() {
          @Override
          public void run() {
            /*
             * firstFuture just completed, but AggregateFuture hasn't yet had time to record the
             * exception in seenExceptions. When we complete secondFuture with the same exception,
             * we want for AggregateFuture to still detect that it's been previously seen.
             */

    bulkFuture.addListener(
        new Runnable() {
          @Override
          public void run() {
            /*
             * firstFuture just completed, but AggregateFuture hasn't yet had time to record the
             * exception in seenExceptions. When we complete secondFuture with the same exception,
             * we want for AggregateFuture to still detect that it's been previously seen.
             */

   *
   * <p>The returned {@link Future} is completed when the pipeline's computation completes, or when
   * the pipeline is cancelled.
   *
   * <p>All objects the pipeline has captured for closing will begin to be closed asynchronously
   * <b>after</b> the returned {@code Future} is done: the future completes before closing starts,
   * rather than once it has finished.
   *

        }
        // re-read ticker now that loading has completed
        long now = map.ticker.read();
        recordRead(e, now);
        return value;
      } finally {
        statsCounter.recordMisses(1);
      }
    }

    @CheckForNull
    V compute(
        K key,
        int hash,

     * than two.)
     *
     * Read operations can thus proceed without locking, but rely on selected uses of volatiles to
     * ensure that completed write operations performed by other threads are noticed. For most
     * purposes, the "count" field, tracking the number of elements, serves as that volatile