Tathagata Das

DeTail: reducing the flow completion time tail in datacenter networks

Web applications have now become so sophisticated that rendering a typical page may require hundreds of intra-datacenter flows. At the same time, web sites must meet strict page creation deadlines of 200-300ms to satisfy user demands for interactivity. Long-tailed flow completion times make it challenging for web sites to meet these constraints. They are forced to choose between rendering a subset of the complex page, or delay its rendering, thus missing deadlines and sacrificing either quality or responsiveness. Either option leads to potential financial loss. In this paper, we present a new cross-layer network stack aimed at reducing the long tail of flow completion times. The approach exploits cross-layer information to reduce packet drops, prioritize latency-sensitive flows, and evenly distribute network load, effectively reducing the long tail of flow completion times. We evaluate our approach through NS-3 based simulation and Click-based implementation demonstrating our ability to consistently reduce the tail across a wide range of workloads. We often achieve reductions of over 50% in 99.9th percentile flow completion times.

Apache Spark: a unified engine for big data processing

This open source computing framework unifies streaming, batch, and interactive big data workloads to unlock new applications.

Scaling spark in the real world: performance and usability

Apache Spark is one of the most widely used open source processing engines for big data, with rich language-integrated APIs and a wide range of libraries. Over the past two years, our group has worked to deploy Spark to a wide range of organizations through consulting relationships as well as our hosted service, Databricks. We describe the main challenges and requirements that appeared in taking Spark to a wide set of users, and usability and performance improvements we have made to the engine in response.

Time-evolving graph processing at scale

Time-evolving graph-structured big data arises naturally in many application domains such as social networks and communication networks. However, existing graph processing systems lack support for efficient computations on dynamic graphs. In this paper, we represent most computations on time evolving graphs into (1) a stream of consistent and resilient graph snapshots, and (2) a small set of operators that manipulate such streams of snapshots. We then introduce GraphTau, a time-evolving graph processing framework built on top of Apache Spark, a widely used distributed dataflow system. GraphTau quickly builds fault-tolerant graph snapshots as each small batch of new data arrives. GraphTau achieves high performance and fault tolerant graph stream processing via a number of optimizations. GraphTau also unifies data streaming and graph streaming processing. Our preliminary evaluations on two representative datasets show promising results. Besides performance benefit, GraphTau API relieves programmers from handling graph snapshot generation, windowing operators and sophisticated differential computation mechanisms.

Structured Streaming: A Declarative API for Real-Time Applications in Apache Spark

With the ubiquity of real-time data, organizations need streaming systems that are scalable, easy to use, and easy to integrate into business applications. Structured Streaming is a new high-level streaming API in Apache Spark based on our experience with Spark Streaming. Structured Streaming differs from other recent streaming APIs, such as Google Dataflow, in two main ways. First, it is a purely declarative API based on automatically incrementalizing a static relational query (expressed using SQL or DataFrames), in contrast to APIs that ask the user to build a DAG of physical operators. Second, Structured Streaming aims to support end-to-end real-time applications that integrate streaming with batch and interactive analysis. We found that this integration was often a key challenge in practice. Structured Streaming achieves high performance via Spark SQLs code generation engine and can outperform Apache Flink by up to 2x and Apache Kafka Streams by 90x. It also offers rich operational features such as rollbacks, code updates, and mixed streaming/batch execution. We describe the systems design and use cases from several hundred production deployments on Databricks, the largest of which process over 1 PB of data per month.