cluster, it will fail, usually with an error message that will direct you here. If you saw that error, then don't worry, the operation succeeded except for the part where the new replication controller is renamed back to the old name. You can just do another rolling update using kubectl 1.4 or higher to change the name back: look for a replication controller that has the original name plus a random suffix.

Unfortunately, there is a much rarer second possible failure mode: the replication controller...

/**
 * An immutable well-formed internet domain name, such as {@code com} or {@code foo.co.uk}. Only
 * syntactic analysis is performed; no DNS lookups or other network interactions take place. Thus
 * there is no guarantee that the domain actually exists on the internet.
 *
 * <p>One common use of this class is to determine whether a given string is likely to represent an

 * into lines. Like {@link BufferedReader}, these methods break lines on any of {@code \n}, {@code
 * \r} or {@code \r\n}, do not include the line separator in each line and do not consider there to
 * be an empty line at the end if the contents are terminated with a line separator.
 *
 * <p>Any {@link ByteSource} containing text encoded with a specific {@linkplain Charset character

 * into lines. Like {@link BufferedReader}, these methods break lines on any of {@code \n}, {@code
 * \r} or {@code \r\n}, do not include the line separator in each line and do not consider there to
 * be an empty line at the end if the contents are terminated with a line separator.
 *
 * <p>Any {@link ByteSource} containing text encoded with a specific {@linkplain Charset character

      assertTrue(afterBurst >= 1000);
    }
  }

  /**
   * This neat test shows that no matter what weights we use in our requests, if we push X amount of
   * permits in a cool state, where X = rate * timeToCoolDown, and we have specified a
   * timeToWarmUp() period, it will cost as the prescribed amount of time. E.g., calling

      assertTrue(afterBurst >= 1000);
    }
  }

  /**
   * This neat test shows that no matter what weights we use in our requests, if we push X amount of
   * permits in a cool state, where X = rate * timeToCoolDown, and we have specified a
   * timeToWarmUp() period, it will cost as the prescribed amount of time. E.g., calling

  // threadpool, so that we avoid the possibility of running the runner_ in the
  // threadpool of GPU event mgr, as that can trigger more callbacks to be
  // scheduled on that same threadpool, causing a deadlock in cases where the
  // caller of event_mgr->ThenExecute() blocks on the completion of the callback
  // (as in the case of ConstOp kernel creation on GPU, which involves copying a
  // CPU tensor to GPU).

        addr = getCachedAddress(name);

        if (addr == null) {
            addr = (NbtAddress) checkLookupTable(name);
            /* This is almost exactly like InetAddress.java. See the
             * comments there for a description of how the LOOKUP_TABLE prevents
             * redundant queries from going out on the wire.
             */
            if (addr == null) {
                try {

			}
			if b != data[i] {
				t.Fatalf("n = %d: incorrect byte returned from ReadByte: got %q, want %q", n, b, data[i])
			}
		}
		// Unread one byte if there is one.
		if n > 0 {
			if err := r.UnreadByte(); err != nil {
				t.Errorf("n = %d: unexpected error on UnreadByte: %v", n, err)
			}
		}
		// Test that we cannot unread any further.

			if errCode != ErrNone {
				if ok {
					tc.FuncName = "handler.Auth"
					tc.ResponseRecorder.LogErrBody = true
				}

				// All our internal APIs are sensitive towards Date
				// header, for all requests where Date header is not
				// present we will reject such clients.
				defer logger.AuditLog(r.Context(), w, r, mustGetClaimsFromToken(r))
				writeErrorResponse(r.Context(), w, errorCodes.ToAPIErr(errCode), r.URL)