Today, big data applications are 10 times more complex than regular applications, and developers often need to understand a lot of technology to make big data work.
The application of big data is still too difficult. Despite the many speculations, most companies are still trying to get value from their data. Dresner Consulting Services concluded: "Despite the long-term awareness raising and hype, the actual deployment of big data analytics is not currently widely applicable to most organizations."
This is a matter of personnel. Despite the persuasive data, corporate executives tend to ignore this data. However, a large part of the complexity of big data is due to the software required. While Spark and other newer systems have improved the trajectory, the big data infrastructure is still too difficult, which is a savvy point for Jesse Anderson.
Difficult to implement
Talent has long been one of the biggest obstacles to big data adoption. In 2015, Bain&Co. found through a survey of senior IT executives that 59% of respondents believe their company lacks the ability to make sense for data and business. Investigator Gartner analyst Nick Hoodco specifically pointed out that "by 2018, due to skills and integration challenges, 70% of Hadoop deployments will not be able to meet cost savings and revenue targets." Staff skills are important, in other words The relevant talents are in short supply.
Over time, the skill gap between people will decrease, of course, but understanding the average Hadoop deployment is not trivial. Anderson pointed out that the complexity of big data boils down to two main factors: "You need to master 10 to 30 different technologies, just to create a big data solution. The use of distributed systems is relatively simple."
How can a big data application complex and diverse?
What's the question
Anderson said the complexity of typical mobile applications and Hadoop-supported applications, noting that the latter involves twice the number of "boxes" or components. However, with simple words, "Helloop's 'Hello World' is more complex than advanced settings in other domains.
Anderson said that people face complex difficulties and need to understand the wide range of systems involved. For example, one might need to know 10 techniques to build a big data application, but this may require familiarizing with another 20 technologies, just knowing which technology to use in a given situation. Otherwise, for example, how would you know to use MongoDB instead of Hbase? Or Cassandra? Or neo4j?
In addition, running in a distributed system has its complexity, and the shortage of big data skills still exists.
Simple way out
One way companies are struggling to minimize the complexity inherent in big data building is to move to the public cloud. According to a recent Databricks survey of Apache Spark users, Spark's deployment to the public cloud has increased by 10% in the past year, reaching 61% of the overall deployment. Cloud computing replaces the cumbersome and inflexible on-premise infrastructure that provides flexibility.
However, it does not eliminate the complexity of the technology involved. The same choice about this or database or message broker still exists. This choice, and the complexity of it, will not disappear soon. Companies like Cloudera and Hortonworks have tried to simplify these choices and integrate them into the stack, but they still basically provide tools that need to be understood to be useful. Amazon Web Services is further evolving through its Lambda service, which enables developers to focus on writing application code, while AWS is responsible for all underlying infrastructure.
But the next step is to pre-create the app for the end user, which is a bigger opportunity for Wall Street analyst Peter Goldmark to sell the infrastructure components. In his words, a major category of “winners†is an application and analytics vendor that abstracts the complexity of the underlying technology into a user-friendly front end. The addressable audience of enterprise users will be more focused on core technologies than the programmer's market.
This is where the market needs to go, and it is fast. People have hardly done it. For every company that can master all the relevant big data technology companies, including those in the high-end industry, just want to just reinvent themselves and need someone to make their data more operative, people now need such suppliers.
What is the Patch Panel? Why is the good choice for your Server Center and Data Center?
A patch panel is a steel panel with a number of ports helps organize a group of cables. Each of those ports includes a wire that goes to a different location. Patch panel could be small size with 2 ports, or very large size with hundreds of ports. They could also be installation for fiber optic cables, RJ45 cables, CAT5E / CAT6 / CAT6A/ CAT7 / CAT8 Ethernet cables, and many others.
CAT6 Patch Panel,patch panel cat6,10 Inch 1U Patch Panel,Home Network Patch Panel
NINGBO UONICORE ELECTRONICS CO., LTD , https://www.uonicore.com