package com.alibaba.jstorm.common.metric.codahale;
import com.codahale.metrics.Reservoir;
import com.codahale.metrics.Snapshot;
import com.codahale.metrics.WeightedSnapshot;
import java.util.ArrayList;
import java.util.concurrent.ConcurrentSkipListMap;
import java.util.concurrent.TimeUnit;
import java.util.concurrent.atomic.AtomicLong;
import java.util.concurrent.locks.ReentrantReadWriteLock;
import static java.lang.Math.exp;
import static java.lang.Math.min;
public class ExponentiallyDecayingReservoir implements Reservoir {
private static final int DEFAULT_SIZE = 1028;
private static final double DEFAULT_ALPHA = 0.015;
private static final long RESCALE_THRESHOLD = TimeUnit.HOURS.toMillis(1);
private final ConcurrentSkipListMap<Double, WeightedSnapshot.WeightedSample> values;
private final ReentrantReadWriteLock lock;
private final double alpha;
private final int size;
private final AtomicLong count;
/**
* start time, epoch seconds
*/
private volatile long startTime;
/**
* next scale time, epoch milliseconds
*/
private final AtomicLong nextScaleTime;
/**
* Creates a new {@link ExponentiallyDecayingReservoir} of 1028 elements, which offers a 99.9%
* confidence level with a 5% margin of error assuming a normal distribution, and an alpha
* factor of 0.015, which heavily biases the reservoir to the past 5 minutes of measurements.
*/
public ExponentiallyDecayingReservoir() {
this(DEFAULT_SIZE, DEFAULT_ALPHA);
}
/**
* Creates a new {@link ExponentiallyDecayingReservoir}.
*
* @param size the number of samples to keep in the sampling reservoir
* @param alpha the exponential decay factor; the higher this is, the more biased the reservoir
* will be towards newer values
*/
public ExponentiallyDecayingReservoir(int size, double alpha) {
this.values = new ConcurrentSkipListMap<>();
this.lock = new ReentrantReadWriteLock();
this.alpha = alpha;
this.size = size;
this.count = new AtomicLong(0);
this.startTime = currentTimeInSeconds();
this.nextScaleTime = new AtomicLong(System.currentTimeMillis() + RESCALE_THRESHOLD);
}
@Override
public int size() {
return (int) min(size, count.get());
}
@Override
public void update(long value) {
update(value, currentTimeInSeconds());
}
/**
* Adds an old value with a fixed timestamp to the reservoir.
*
* @param value the value to be added
* @param timestamp the epoch timestamp of {@code value} in seconds
*/
public void update(long value, long timestamp) {
rescaleIfNeeded();
lockForRegularUsage();
try {
final double itemWeight = weight(timestamp - startTime);
final WeightedSnapshot.WeightedSample sample = new WeightedSnapshot.WeightedSample(value, itemWeight);
final double priority = itemWeight / ThreadLocalRandom.current().nextDouble();
final long newCount = count.incrementAndGet();
if (newCount <= size) {
values.put(priority, sample);
} else {
Double first = values.firstKey();
if (first < priority && values.putIfAbsent(priority, sample) == null) {
// ensure we always remove an item
while (values.remove(first) == null) {
first = values.firstKey();
}
}
}
} finally {
unlockForRegularUsage();
}
}
private void rescaleIfNeeded() {
final long now = System.currentTimeMillis();
final long next = nextScaleTime.get();
if (now >= next) {
rescale(now, next);
}
}
@Override
public Snapshot getSnapshot() {
lockForRegularUsage();
try {
return new WeightedSnapshot(values.values());
} finally {
unlockForRegularUsage();
}
}
private long currentTimeInSeconds() {
return TimeUnit.MILLISECONDS.toSeconds(System.currentTimeMillis());
}
private double weight(long t) {
return exp(alpha * t);
}
/* "A common feature of the above techniques—indeed, the key technique that
* allows us to track the decayed weights efficiently—is that they maintain
* counts and other quantities based on g(ti − L), and only scale by g(t − L)
* at query time. But while g(ti −L)/g(t−L) is guaranteed to lie between zero
* and one, the intermediate values of g(ti − L) could become very large. For
* polynomial functions, these values should not grow too large, and should be
* effectively represented in practice by floating point values without loss of
* precision. For exponential functions, these values could grow quite large as
* new values of (ti − L) become large, and potentially exceed the capacity of
* common floating point types. However, since the values stored by the
* algorithms are linear combinations of g values (scaled sums), they can be
* rescaled relative to a new landmark. That is, by the analysis of exponential
* decay in Section III-A, the choice of L does not affect the final result. We
* can therefore multiply each value based on L by a factor of exp(−α(L′ − L)),
* and obtain the correct value as if we had instead computed relative to a new
* landmark L′ (and then use this new L′ at query time). This can be done with
* a linear pass over whatever data structure is being used."
*/
private void rescale(long now, long next) {
if (nextScaleTime.compareAndSet(next, now + RESCALE_THRESHOLD)) {
lockForRescale();
try {
final long oldStartTime = startTime;
this.startTime = currentTimeInSeconds();
final double scalingFactor = exp(-alpha * (startTime - oldStartTime));
final ArrayList<Double> keys = new ArrayList<>(values.keySet());
for (Double key : keys) {
final WeightedSnapshot.WeightedSample sample = values.remove(key);
final WeightedSnapshot.WeightedSample newSample =
new WeightedSnapshot.WeightedSample(sample.value, sample.weight * scalingFactor);
values.put(key * scalingFactor, newSample);
}
// make sure the counter is in sync with the number of stored samples.
count.set(values.size());
} finally {
unlockForRescale();
}
}
}
private void unlockForRescale() {
lock.writeLock().unlock();
}
private void lockForRescale() {
lock.writeLock().lock();
}
private void lockForRegularUsage() {
lock.readLock().lock();
}
private void unlockForRegularUsage() {
lock.readLock().unlock();
}
}