TagsArticle Source
Apache Hadoop vs. IBM Platform Symphony and Infosphere BigInsights
IBM has completed several big data benchmarks of significance employing IBM Platform Symphony and various Hadoop distributions including IBM Infosphere BigInsights. Platform Symphony is a distributed computing and big data analytics product widely used in large scale grid computing environments. IBM InfoSphere BigInsights brings the power of Hadoop to the enterprise. Organizations using the two products together get the benefit of a multi-tenant, heterogeneous application cluster with higher utilization and performance. Using InfoSphere BigInsights you can gain new insights from a combination of data sources and overcome the high costs of converting unstructured data sources to a structured format.
Benchmarks that IBM has run include:
- The Terasort benchmark, run on an IBM cloud using IBM InfoSphere BigInsights. In an unaudited result on a 1,000 Virtual Machine IBM private cloud environment, IBM obtained a breakthrough Terasort result, sorting 100 terabytes in 10,369 seconds, slightly less than 3 hours. This bettered a prior world-record result, but required only 10% of the cores used in setting the previous record.
- A Contrail bio workload, an open-source software package used for de novo genome assembly. In a controlled test conducted in March of 2013, Platform Symphony Advanced Edition was found to reduce the time required to sequence a 10K read sample from a reference e-coli bacteria genome by a factor of 3.4 on an eight-node Apache Hadoop cluster. Read the report here.
- The SWIM benchmark (Statistical Workload Injector for MapReduce), a benchmark representing a real-world big data workload developed by University of California at Berkley in close cooperation with Facebook. This test provides rigorous measurements of the performance of MapReduce systems comprised of real industry workloads. In an audited result conducted by STAC Research, InfoSphere BigInsights, powered by Platform Symphony technology, accelerated SWIM/Facebook workload traces run using open-source Apache Hadoop by an average of approximately 4 times. A STAC Report featuring this result published in October of 2013 is available for download here.
- The “sleep” benchmark, which has been used as a test of the efficiency of core scheduling of MapReduce workloads and which was promoted at Hadoop World 2011. IBM demonstrated that Platform Symphony Advanced Edition runs the sleep test almost 11 times faster than Apache Hadoop alone, demonstrating that low-latency scheduling is critical to maximizing Hadoop workload performance.
Audited Report STAC Report - UCB SWIM & Hadoop
Benchmark Faster Genome sequencing with contrail-bio
Whitepaper Analyzing data at the speed of business
Get the Platform Symphony Advantage
While a major benefit of IBM Platform Symphony is its ability to support diverse applications in a multi-tenant environment while ensuring service levels, these performance tests show that IBM Platform Symphony also helps provide dramatically better performance and efficiency, as well as superior management and monitoring.
Clients using this technology in conjunction with Infosphere BigInsights can get a fully supported high performance Hadoop stack with ease of use, higher productivity with built-in accelerators, and management tools.
- Deliver faster and more accurate analysis for big data applications by doing greater processing with less infrastructure
- Lower costs through reduction in infrastructure and administration overhead
- Enable business agility by supporting multiple groups and diverse workloads on a single shared cluster
These results are important not only because they demonstrate faster MapReduce job execution times, but because they show that organizations running Hadoop workloads can save a significant amount of money on computing infrastructure by using IBM Platform Symphony.
Terasort Benchmark
Running IBM InfoSphere BigInsights on a private cloud environment managed by IBM Platform Symphony in August of 2012, IBM demonstrated a 100 TB terasort result on a cluster comprised of 1,000 virtual machines, 200 physical nodes and 2,400 processing cores. Running the industry standard Terasort benchmark in this private cloud, IBM beat a prior world-record^1^ using 17 times fewer servers and 12 times fewer total processing cores. This result showed not only that it is straightforward to build a large-scale Hadoop environment using IBM’s cloud-based solutions, but that big data workloads with IBM BigInsights can be run more economically using IBM Platform Symphony, providing dramatic savings related to infrastructure, power and facilities.
Hardware
- 200 IBM dx360M3 computers in iDataPlex racks
- 2 IBM dx360M3 computers in iDataPlex racks as master hosts
- 120 GB memory per host, 12 x 3 TB spindles per host
- 2,400 cores
Software
- 1000 Virtual machines
- RHEL 6.2 with KVM
- IBM InfoSphere BigInsights 1.3.0.1
- IBM Platform Symphony Advanced Edition 5.2
- IBM Platform Symphony BigInsights Integration Path for 1.3.0.1
Contrail-bio Genome Sequencing Benchmark
Contrail is an open-source software effort that leverages Hadoop MapReduce to accelerate de novo genome assembly. During March of 2013, IBM conducted a series of tests to understand the performance advantage that Symphony could offer on a reference Hadoop cluster running a 10K read sample of an e-coli bacterium included as part of the Contrail software suite. In an eight-node Hadoop cluster with 108 cores dedicated to Map and Reduce tasks, Platform Symphony was found to compute results 3.4 times faster than Hadoop alone, reducing the job run-time from 873 seconds to 258 seconds on the same cluster and dataset. Get the results
SWIM Benchmark
Equally compelling are results obtained using social media workloads. The SWIM benchmark developed at University of California, Berkeley with co-operation from Facebook, measures real-world MapReduce workloads by simulating traces of application activity captured at Facebook in 2009 and 2010. The benchmark authors view it as a rigorous predictor of MapReduce performance^2^.
Using this benchmark, IBM demonstrated, in results audited by an independent testing organization, that by augmenting a Hadoop cluster with BigInsights powered by Platform Symphony technology, the simulated Facebook workloads were speeded up by a mean of 3.8X compared with running on Apache Hadoop alone. As a corollary, given the nature of the SWIM benchmark, this result demonstrated that equivalent performance with Symphony could have been obtained with dramatically less hardware and lower infrastructure cost.
Hardware
- 17 IBM System x3630M3 computers as compute nodes
- 1 IBM System x3630M3 as master host
Software
- Apache Hadoop 1.1.2
- IBM InfoSphere BigInsights Enterprise Edition 2.1.0.1 powered by Adaptive MapReduce^3^
Results
- Mean speedup of BigInsights versus Apache Hadoop on 302 jobs of SWIM benchmark Facebook workload: 3.8X
Sleep Benchmark
The Hadoop “sleep” benchmark shared at Hadoop World in 2011^4^ was run to demonstrate the relative scheduling efficiency of IBM Platform Symphony to competing Hadoop distributions. This was an audited result published by a third party. Running this standard test, which is promoted as a measure of scheduling efficiency, IBM InfoSphere BigInsights powered by Platform Symphony technology achieved a speedup of just under 11X for a Hadoop 1.1.2 sleep test result comprised of 5000 x 1 msec map tasks.
Hardware
- 17 IBM System x3630M3 computers as compute nodes
- 1 IBM System x3630M3 as master host
Software
- Apache Hadoop 1.1.2
- InfoSphere BigInsights Enterprise Edition 2.1.0.1 powered by Adaptive MapReduce
Results
- Mean task throughput 5,000 map tasks on BigInsights – 665.26 tasks / second
- Task throughput 5,000 map tasks with Hadoop – 61.68 tasks / second
IBM Platform Symphony brings many advantages to distributed computing environments including multi-tenancy, guaranteed service levels, superior management tools, and support for diverse, heterogeneous workloads. These benchmark results demonstrate that IBM Platform Symphony can provide dramatic performance advantages and financial savings to customers deploying big data environments. For IBM InfoSphere BigInsights users, or those considering open-source or derivative Hadoop environments, IBM Platform Symphony can help accelerate Hadoop workloads while reducing cost and improving workload reliability.
The result established by Yahoo in 2009 is regarded by many as a de-facto standard 100 TB sorts result. Clearly technology changed in three years, but what is notable is that this 2012 result was achieved running in VMs while the Yahoo result was achieved on bare-metal, faster in theory. SWIM benchmark details
Adaptive MapReduce is based on a low latency MapReduce implementation provided by IBM Platform Symphony that is included as a core feature in IBM BigInsights Enterprise Edition. Benchmark presentation on Slideshare