com.netflix.spectator.api.histogram

Class BucketFunctions

java.lang.Object

com.netflix.spectator.api.histogram.BucketFunctions

public final class BucketFunctions
extends java.lang.Object

Helpers for creating bucketing functions.

Method Summary

Modifier and Type	Method and Description
static java.util.function.LongFunction<java.lang.String>	age(long max, java.util.concurrent.TimeUnit unit) Returns a function that maps age values to a set of buckets.
static java.util.function.LongFunction<java.lang.String>	ageBiasOld(long max, java.util.concurrent.TimeUnit unit) Returns a function that maps age values to a set of buckets.
static java.util.function.LongFunction<java.lang.String>	bytes(long max) Returns a function that maps size values in bytes to a set of buckets.
static java.util.function.LongFunction<java.lang.String>	decimal(long max) Returns a function that maps size values to a set of buckets.
static java.util.function.LongFunction<java.lang.String>	latency(long max, java.util.concurrent.TimeUnit unit) Returns a function that maps latencies to a set of buckets.
static java.util.function.LongFunction<java.lang.String>	latencyBiasSlow(long max, java.util.concurrent.TimeUnit unit) Returns a function that maps latencies to a set of buckets.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Detail

age

public static java.util.function.LongFunction<java.lang.String> age(long max,
                                                                    java.util.concurrent.TimeUnit unit)

Returns a function that maps age values to a set of buckets. Example use-case would be tracking the age of data flowing through a processing pipeline. Values that are less than 0 will be marked as "future". These typically occur due to minor variations in the clocks across nodes. In addition to a bucket at the max, it will create buckets at max / 2, max / 4, and max / 8.

Parameters:

max - Maximum expected age of data flowing through. Values greater than this max will be mapped to an "old" bucket.

unit - Unit for the max value.

Returns:

Function mapping age values to string labels. The labels for buckets will sort so they can be used with a simple group by.

latency

public static java.util.function.LongFunction<java.lang.String> latency(long max,
                                                                        java.util.concurrent.TimeUnit unit)

Returns a function that maps latencies to a set of buckets. Example use-case would be tracking the amount of time to process a request on a server. Values that are less than 0 will be marked as "negative_latency". These typically occur due to minor variations in the clocks, e.g., using System.currentTimeMillis() to measure the latency and having a time adjustment between the start and end. In addition to a bucket at the max, it will create buckets at max / 2, max / 4, and max / 8.

Parameters:

max - Maximum expected age of data flowing through. Values greater than this max will be mapped to an "old" bucket.

unit - Unit for the max value.

Returns:

Function mapping age values to string labels. The labels for buckets will sort so they can be used with a simple group by.

ageBiasOld

public static java.util.function.LongFunction<java.lang.String> ageBiasOld(long max,
                                                                           java.util.concurrent.TimeUnit unit)

Returns a function that maps age values to a set of buckets. Example use-case would be tracking the age of data flowing through a processing pipeline. Values that are less than 0 will be marked as "future". These typically occur due to minor variations in the clocks across nodes. In addition to a bucket at the max, it will create buckets at max - max / 8, max - max / 4, and max - max / 2.

Parameters:

max - Maximum expected age of data flowing through. Values greater than this max will be mapped to an "old" bucket.

unit - Unit for the max value.

Returns:

Function mapping age values to string labels. The labels for buckets will sort so they can be used with a simple group by.

latencyBiasSlow

public static java.util.function.LongFunction<java.lang.String> latencyBiasSlow(long max,
                                                                                java.util.concurrent.TimeUnit unit)

Returns a function that maps latencies to a set of buckets. Example use-case would be tracking the amount of time to process a request on a server. Values that are less than 0 will be marked as "negative_latency". These typically occur due to minor variations in the clocks, e.g., using System.currentTimeMillis() to measure the latency and having a time adjustment between the start and end. In addition to a bucket at the max, it will create buckets at max - max / 8, max - max / 4, and max - max / 2.

Parameters:

max - Maximum expected age of data flowing through. Values greater than this max will be mapped to an "old" bucket.

unit - Unit for the max value.

Returns:

Function mapping age values to string labels. The labels for buckets will sort so they can be used with a simple group by.

bytes

public static java.util.function.LongFunction<java.lang.String> bytes(long max)

Returns a function that maps size values in bytes to a set of buckets. The buckets will use binary prefixes representing powers of 1024. If you want powers of 1000, then see decimal(long).

Parameters:

max - Maximum expected size of data being recorded. Values greater than this amount will be mapped to a "large" bucket.

Returns:

Function mapping size values to string labels. The labels for buckets will sort so they can be used with a simple group by.

decimal

public static java.util.function.LongFunction<java.lang.String> decimal(long max)

Returns a function that maps size values to a set of buckets. The buckets will use decimal prefixes representing powers of 1000. If you are measuring quantities in bytes and want powers of 1024, then see bytes(long).

Parameters:

max - Maximum expected size of data being recorded. Values greater than this amount will be mapped to a "large" bucket.

Returns:

Function mapping size values to string labels. The labels for buckets will sort so they can be used with a simple group by.