Large Data Hadoop Solution

According to the latest Forrest report, many companies are trying to tap into the vast amounts of data they have, including structured, unstructured, semi-structured, and binary data, and explore the use of large data. The following are some of the conclusions of the report:

Most companies estimate that they only analyze 12% of the existing data and that the remaining 88% have not been fully exploited. The lack of a large number of data islands and analytical capabilities is the main cause of this situation. Another problem is how to judge whether data is valuable or not. Especially in the big data age, you have to collect and store this data. Some seemingly unrelated data, such as mobile phone GPS data, could also be a gold mine in the future.

So, a lot of companies are pinning their hopes on using Hadoop to solve the following challenges:

Collect and store all data related to the business functions of the company. Support advanced analytical functions, including business intelligence, using modern methods for advanced visualization and predictive analysis of data. Quickly share your data with the people you need. Consolidating multiple data islands helps organizations solve complex problems that no one has ever mentioned before, even unknown. accommodate the growing volume of data and new data sources. Hadoop supports the rapid and effective expansion of the solution scale, enabling ever-increasing capacity, speed, and diverse data to be controlled by the company.

According to the report, the purchasing cycle of Hadoop is on the rise, so more and more manufacturers are being created in this field. Although Hadoop is an open source project for Apache, anyone can download it for free, most consumers prefer to adopt a vendor's packaging scheme. In addition to packaging all the Hadoop components and ensuring their proper use (compatible versions), vendors typically provide enterprise-class support and extension: Apache Hadoop (common,hdfs,mapreduce) as the core component of the solution, With extra functionality to enhance Hadoop, and add differentiation to make the solution more appealing.

The Forrester report details 9 vendors: Amazon Web Services, Cloudera, Hortonworks, IBM, Intel, MAPR Technology, Microsoft, pivotal software and Teradata, These vendors are evaluated based on the following metrics:

Existing products, including solution architecture, data and processing capabilities, installation, management, monitoring tools, compatibility, and community maturity. Strategies, including vendors ' plans to meet current customer needs and make up for enterprise deployments. The evaluation of the strategy includes equity mergers, the implementation of internal strategies, product roadmap and customer support capabilities. Market share, including the company's financial, global share, installation base, and other software vendors, professional services vendors and software as a service (SAAS)/cloud/hosting provider between the strategic partnership.

The main findings of the report are as follows:

Manufacturers are numerous, but do not dominate the manufacturer.

In this large data Hadoop solution evaluation, the leading vendors are Amazon Web Services, Cloudera, Hortonworks, IBM, MAPR Technology, Pivotal software and Teradata. These vendors are based on the Apache Open Source project, and then add packaging, support, integration and other features, as well as their own innovation and other content to make up the Hadoop in the Enterprise short board. All vendors have implemented these functions, albeit in a slightly different way – as evidenced by the evaluation scores of vendors and vendor data.

The solution of the new-Jin manufacturers also has a decent performance.

In this large data Hadoop solution evaluation, the strong performance of the vendors are Intel and Microsoft. Microsoft has developed a strong roadmap for hdinsight products, making it as competitive as other leading vendors. Microsoft Hdinsight has special optimizations for Azure, so it's the best solution for Microsoft customers who want to implement Hadoop on Azure. While Intel focuses its innovation capabilities on the chip level, it needs to strengthen its strategy and enterprise tools if it wants to make its products an enterprise-class solution.

According to the report, although this review of the large data Hadoop solution market is only the beginning, the report provides good reference information for companies struggling to get out of the complex Hadoop business scenario.