com.thesett.common.throttle.BatchedThrottle Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of junit-toolkit Show documentation
Show all versions of junit-toolkit Show documentation
JUnit Toolkit enhances JUnit with performance testing, asymptotic behaviour analysis, and concurrency testing.
/*
* Copyright The Sett Ltd, 2005 to 2014.
*
* Licensed 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 com.thesett.common.throttle;
/**
* BatchedThrottle is a {@link SleepThrottle} that uses batching to achieve much higher throttling rates than a sleep
* throttle can. Sleep throttle has difficulties once the rate gets above a few hundred hertz, because the JVM cannot
* generate timed pauses that are that short. BatchedThrottle gets around this by only inserting pauses once every so
* many calls to the {@link Throttle#throttle()} method, and using a sleep throttle run at a lower rate. The rate for
* the sleep throttle is chosen so that it remains under 100hz. The final throttling rate of this throttle is equal to
* the batch size times the rate of the underlying sleep throttle.
*
* The batching calculation involves taking the log to the base 100 of the desired rate and rounding this to an
* integer. The batch size is always an exact power of 100 because of the rounding. The rate for an underlying sleep
* throttle is then chosen appropriately.
*
* In practice, the accuracy of a BacthedThrottle skews off but can sometimes even be reasonable up to ten thousand
* hertz compared with 100 Hz for a {@link SleepThrottle}.
*
*
CRC Card
* Responsibilities Collaborations
* Accept throttling rate in operations per second.
* Inject short pauses, occasionaly, to fill out processing cycles to a specified rate.
* Check against a throttle speed without waiting.
*
*
* @author Rupert Smith
* @todo Should always round the log base 100 down to the nearest integer?
*/
public class BatchedThrottle extends BaseThrottle
{
/** Holds the batch size. */
int batchSize;
/** The call count within the current batch. */
long callCount;
/** Holds a sleep throttle configured to run at the batched rate. */
private final Throttle batchRateThrottle = new SleepThrottle();
/**
* Specifies the throttling rate in operations per second.
*
* @param hertz The throttling rate in cycles per second.
*/
public void setRate(float hertz)
{
// Pass the rate unaltered down to the base implementation, for the check method.
super.setRate(hertz);
// Log base 10 over 2 is used here to get a feel for what power of 100 the total rate is.
// As the total rate goes up the powers of 100 the batch size goes up by powers of 100 to keep the
// throttle rate in the range 1 to 100.
int x = (int) (Math.log10(hertz) / 2);
batchSize = (int) Math.pow(100, x);
float throttleRate = hertz / batchSize;
// Reset the call count.
callCount = 0;
// Set the sleep throttle wrapped implementation at a rate within its abilities.
batchRateThrottle.setRate(throttleRate);
}
/**
* Throttle calls to this method to the rate specified by the {@link #setRate(float)} method.
*
* @throws InterruptedException If interrupted whilst performing a blocking wait on the throttle.
*/
public void throttle() throws InterruptedException
{
if ((callCount++ % batchSize) == 0)
{
batchRateThrottle.throttle();
}
}
}
© 2015 - 2024 Weber Informatics LLC | Privacy Policy