Cloud services new Darling Spark and Hadoop, who will be the last winner

Spark first issued by Databricks, financing 33 million dollars; Hadoop is again mapr $110 million trillion in financing to boost its growth in the fierce market competition. In the future large data processing, spark will simplify the existing pipeline processing, integration of a variety of functions, making data processing faster, more convenient and more flexible; Hadoop will also read and write large data in a faster, simpler way. The huge amount of funding will boost the development of spark and Hadoop, and how will they stand up to the big data market in the future? Let's wait and see.

Databricks released Spark, financing 33 million dollars

Databricks recently released spark cloud services at the Spark summit and raised $33 million trillion in venture capital. The company's new Spark cloud service employs the spark framework, which is said to be faster, more convenient, and more flexible, and its cloud services are designed to simplify existing data pipeline processing, include a large number of data storage and data-processing systems that require enterprise management, and many such as to establish and display the machine learning model of the various processing engine, "notepad" and dashboard functions, and so on effective integration.

The vendor of Hadoop MAPR, considering the large amount of capital and human resources invested in MapReduce's technology and ecosystems, hesitates to spark's current application, but still supports the processing framework and believes it will be faster and simpler to read and write large data than MapReduce.

MAPR refinancing 110 million dollars to drive Hadoop growth

At the same time, MAPR also raised 110 million of dollars to drive the development of Hadoop. The CEO of the company affirms the company's advantage over proprietary tools, providing customers with more choices in components: MAPR supports a handful of sql-on-hadoop tools, including hive and drill, and also supports cloudera-developed Impala And HP's Vertica software.

The MAPR's financing in the Hadoop sector has increased markedly, with total financing exceeding 59 million U.S. dollars. Of course, for Hadoop financing, Cloudera's 5 rounds of financing amounted to $300 million trillion, hortonwork from the start of the first to 2013 3 years of financing amount to 198 million dollars.

Where Spark and Hadoop are heading

In the fierce market competition, the future development of spark and Hadoop is more or less disputed.

The data for the Spark service is stored by default in Amazon S3, and if the user has a Hadoop cluster that already works on AWS, it can also store the data in HDFs. Databricks can read data from MongoDB, MySQL, and Amazon redshift, or export data. The company is expected to support a hybrid cloud-local spark environment, placed in a fully open spark cloud, to efficiently implement workload portability.

Hadoop's financing Enterprise MAPR chose Google as one of its partners, and their direction represents the future of big data. Hadoop vendor competitors also face huge competition challenges.

As an effective data-processing framework, spark and Hadoop each has obvious advantages, as users, faster and more convenient to solve their own needs may be a factor in their choice, the financiers should strengthen their service application function, better serve the customer, in order to bring considerable market profits.

Original link: http://www.csdn.net/article/2014-07-02/2820494