Skip to main content

Twitter's 100th open source GitHub repo - Summingbird


With the open sourcing of Twitter’s Summingbird library this week, we now have another ingenous option to integrate real-time and batch processing. Summingbird allows building MapReduce pipelines with simple primitives and merging data sets with low latency.

Working as an abstraction layer, it aims to separate the MapReduce computation and workflow from physical system layer. Within Twitter it has been used to deploy MapReduce workflow on Hadoop and Storm clusters. More specifically, the team of Oscar Boykin, Sam Ritchie and Ashutosh Singhal have been using this library to write MapReduce “streaming computation once and execute it in batch-mode on Scalding, in realtime mode on Storm, or on both Scalding and Storm in a hybrid batch/realtime mode.”

Summingbird’s real-time MapReduce DSL takes advantage of 3 computation libraries from Twitter:
•    Algebird: Abstract algebra library for Scala targeted at building aggregation systems
•    Bijection: An invertible function that converts back and forth between two types, with the contract that a round-trip through the Bijection will bring back the original object.
•    Storehaus: A library built on top of Twitter's Future that makes it easy to work with asynchronous key value stores.
The library leverages monoids in particular for advanced statistical analysis. Within Algebird, these find application for Bloom filters, HyperLogLog counters and count-min sketches. A monoid is a set that is closed under an associative binary operation and has an identity element I (element of) S such that for all a (element of) S, I a = a I = a.


Summingbird has been used to power Twitter headlines which shows stories related to tweets. It is also believed to be a key driver behind providing near real time views to advertisers on the Twitter advertisement dashboard. Another key milestone that the project would mark is that it shall be Twitter’s 100th public open source code repository on GitHub.

You may follow @summingbird for more updates. The source code has been released this week on https://github.com/twitter/summingbird.


------------------------------
update: 4 Sep 2013: text updated to reflect authors and link after the project release.

image source: Sam Ritchie/Twitter

Comments

Popular posts from this blog

Offloading legacy with Hadoop

With most Fortune 500 organizations having invested in mainframes and other workload systems in the past, the rise of Big Data platforms poses newer integration challenges. The data integration and ETL players are finding fresh opportunities to solve business and IT problems within the Hadoop ecosystem.
To understand the context, challenges and opportunities, we asked a few questions to Syncsort CEO Lonne Jaffe. Syncsort provides fast, secure, enterprise-grade software spanning Big Data in Apache Hadoop to Big Iron on mainframes. At Syncsort, Lonne Jaffe is focusing on accelerating the growth of the company's high-performance Big Data offerings, both organically and through acquisition.
From mainframes to Hadoop and other platforms, Syncsort seems to have been evolving itself continuously. Where do you see Syncsort heading further?Lonne Jaffe: Syncsort is extraordinary in its ability to continuously reinvent itself. Today, we’re innovating around Apache Hadoop and other Big Data pla…

Data deduplication tactics with HDFS and MapReduce

As the amount of data continues to grow exponentially, there has been increased focus on stored data reduction methods. Data compression, single instance store and data deduplication are among the common techniques employed for stored data reduction.
Deduplication often refers to elimination of redundant subfiles (also known as chunks, blocks, or extents). Unlike compression, data is not changed and eliminates storage capacity for identical data. Data deduplication offers significant advantage in terms of reduction in storage, network bandwidth and promises increased scalability.
From a simplistic use case perspective, we can see application in removing duplicates in Call Detail Record (CDR) for a Telecom carrier. Similarly, we may apply the technique to optimize on network traffic carrying the same data packets.
Some of the common methods for data deduplication in storage architecture include hashing, binary comparison and delta differencing. In this post, we focus on how MapReduce and…

Top Big Data Influencers of 2015

2015 was an exciting year for big data and hadoop ecosystem. We saw hadoop becoming an essential part of data management strategy of almost all major enterprise organizations. There is cut throat competition among IT vendors now to help realize the vision of data hub, data lake and data warehouse with Hadoop and Spark.
As part of its annual assessment of big data and hadoop ecosystem, HadoopSphere publishes a list of top big data influencers each year. The list is derived based on a scientific methodology which involves assessing various parameters in each category of influencers. HadoopSphere Top Big Data Influencers list reflects the people, products, organizations and portals that exercised the most influence on big data and ecosystem in a particular year. The influencers have been listed in the following categories:

AnalystsSocial MediaOnline MediaProductsTechiesCoachThought LeadersClick here to read the methodology used.

Analysts:Doug HenschenIt might have been hard to miss Doug…