Github user tdas commented on a diff in the pull request:

    https://github.com/apache/spark/pull/7471#discussion_r35048518
  
    --- Diff: 
streaming/src/test/scala/org/apache/spark/streaming/receiver/RateLimiterSuite.scala
 ---
    @@ -0,0 +1,47 @@
    +/*
    + * Licensed to the Apache Software Foundation (ASF) under one or more
    + * contributor license agreements.  See the NOTICE file distributed with
    + * this work for additional information regarding copyright ownership.
    + * The ASF licenses this file to You under the Apache License, Version 2.0
    + * (the "License"); you may not use this file except in compliance with
    + * the License.  You may obtain a copy of the License at
    + *
    + *    http://www.apache.org/licenses/LICENSE-2.0
    + *
    + * Unless required by applicable law or agreed to in writing, software
    + * distributed under the License is distributed on an "AS IS" BASIS,
    + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
    + * See the License for the specific language governing permissions and
    + * limitations under the License.
    + */
    +
    +package org.apache.spark.streaming.receiver
    +
    +import org.apache.spark.SparkConf
    +import org.apache.spark.SparkFunSuite
    +
    +/** Testsuite for testing the network receiver behavior */
    +class RateLimiterSuite extends SparkFunSuite {
    +
    +  test("rate limiter initializes even without a maxRate set") {
    --- End diff --
    
    I was not sure whether the Guava rate limiter behaves in the desired way 
when the rate is updated. There are several ways of implementing a updateable 
rate limiter, and I wanted to understand this's one behavior, and cover it in 
tests. However, I just digged deeper in Guava RateLimiter code, and realized 
that there are fundamentally bigger problems in using the Guava rate limiter, 
even for stable rate. It works effectively using token bucket philosophy, which 
allows a large burst of data to be allowed, if there has been no data for 
while. We DONT want that. This means that we need to implement our own rate 
limiter (which was the case, until someone replaced my implementation with 
Guava). Anyways, thats not something outside the scope of this PR. So I am okay 
not testing the rate limiting. 
    
    So please revert ReceiverSuite, and remove the block manager unit tests in 
this testsuite. Basically, revert it to the previous iteration (sorry :) )
    - 


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to