Distributed Data Generator for AdImpression Events. (by default) we generate 100 campaigns, with 10 ads each. We write those 1000 ads into Redis, wi

Distributed Data Generator for AdImpression Events. Designed to generate large numbers of campaigns

Generate timestamp and watermarks for data stream

Generate timestamp and watermarks

The states of address detection mechanism. There is only a state transition if the current state failed to determine the address.

This is a basic example of a Storm topology.

To Run: flink run -c flink.benchmark.state.AdvertisingTopologyFlinkState target/flink-benchmarks-0.1.0.jar "../conf/benchmarkConf.yaml" <p> <p> Imple

To Run: flink run -c flink.benchmark.state.AdvertisingTopologyFlinkStateHighKeyCard target/flink-benchmarks-0.1.0.jar "../conf/benchmarkConf.yaml" <p

To Run: flink run -c flink.benchmark.AdvertisingTopologyFlinkWindows target/flink-benchmarks-0.1.0.jar "../conf/benchmarkConf.yaml" This job variant

This is a basic example of a Storm topology.

To Run: flink run -c flink.benchmark.AdvertisingTopologyNative target/flink-benchmarks-0.1.0.jar "../conf/benchmarkConf.yaml Computes the windows us

To Run: flink run -c flink.benchmark.AdvertisingTopologyRedisDirect target/flink-benchmarks-0.1.0.jar "../conf/benchmarkConf.yaml"

Encapsulating configuration in once place

Write windows to Redis

Build windows directly in Redis

Utilities to determine the network interface and address that should be used to bind the TaskManager communication to.

Parse JSON

Parse JSON

Deserialize the JSON

Deserialize JSON

Parse JSON

Custom trigger - Fire and purge window when window closes, also fire every 1000 ms.

Filter out everything except "view" events

Filter out all but "view" events

Filter down to only "view" events

Filter out all but "view" events.

A data generator source

A data generator for generating large numbers of campaigns

Simple util to print kafka partitions locally

Base class for LoadGenerator Sources. Implements features to generate whatever load you require.

Project to campaign id

Custom operator that computes windows and also makes that state directly queryable from outside the streaming system via Akka.

Some basic Redis support for preparing to run tests compatible with Yahoo's original benchmarking tools.

Map ad ids to campaigns using cached data from Redis

Map from ad id to campaign using cached Redis data

Map from ad to campaign using Redis

Sink computed windows to Redis

Simplified version of Redis data structure

Write a string into Kafka as fast as possible