The challenge in building a BI system and Big data analyticsThe solutionAn integrated BI platform and Big Data analytics is a different system. You have the right to choose whether to buy or build. You must review the existing system, the case of use, level of experience and competence of your staff. Some companies may want to build a system using only the open source Hadoop (Hadoop Distributed File System [HDFS] and that's MapReduce), Zookeeper, Solr, Sqoop, Hive, HBase, Nagios, and Cacti, while others may seek more support and trying to build a system using IBM ® BigInsights ® ™ and IBM InfoSphere Netezza. Other companies may want to split the data are structured and unstructured, and build an interface (GUI) graphical user class for normal users, users have many powers, and other applications.It really depends on the company. And it is not just a system plug-and-play. Although your decision to buy or build at each level has different parts.ETLETL, implementation stages and control data, and all related process is always an important first step. You can not put the Big Application Data on a transaction system and expect that everything works without compromising the original system, or expect it integrates well with everything that the system still works. Therefore, some data need to be brought into the Hadoop or any other noSQL system or a Data Warehouse parallel processing (MPP). There are many tools and methods to do this, and most of them depend on the system, source code, data, the size and manpower.You can start with Sqoop. It is a great tool to process data from the management system relational database. Add the other open source tools such as the Flume or Scribe have support writing log. There are also other tools such as the IBM InfoSphere DataStage or Talend ETL ®, both of which have integrated the Big Data. These tools more intuitive and do not need to have a PhD in computers to building infrastructure. Both tools provide technical documentation, updates, and intuitive interface, we are always improving, and are used in many industries and in enterprises.Some companies like to use open source. Other companies may have many systems built on IBM products. Clearly, what integration has to be used with the new technology is important to consider.It is time you spend to build ETL system, and sadly if results are not as you expect. Hadoop has many components that you may need to than Sqoop. The integrated and complementary elements can cause side effects, especially if you have no experience and knowledge or want to manually build ETL tool. This process requires time and patience. Maybe you will also encounter many obstacles. You can use an open source tool for the community. Or you can configure and develop its own ETL tools with internal applications and open-source tool, and then, if the open source community have the right to change or a few of your development staff no longer working anymore, just at the moment you will have a system that nobody knew how to maintain or repair.The wise business focused on the staff, experience, budget, the potential and the reality of them. For example, if a business has a team relatively small IT staff, the comparison of building system with Google or Facebook is not a good idea. Don't ever compare your small company with the company have available server systems and computer experts working on the system and specific infrastructure. Sometimes, using cloud services or external staff might be the only option. The other times, the Big device Data such as Netezza is the best choice.ArchiveData storage is a huge factor and may request that you use many different technologies. In the system of Hadoop, HBase. But some companies use Neo4j, Cassandra, Netezza, HDFS, and other technologies, depending on what is needed. HDFS is a file storage system. HBase is a system hosted by column (column) similar to Cassandra. Many companies use Cassandra for the analysis closer to real time. However HBase also are increasingly being developed.You can look between HBase or Cassandra when wanting to use a management system open source database for the analysis of Big Data. According to the Data Warehouse platform, Netezza is one of the leading technology in the BI and analytical technology. The best choice to integrate Big Data is using an integrated platform including Hadoop and Cassandra for unstructured data or semi-structured and structured data for Netezza.IBM Customer Intelligence Netezza Appliance that combines a number of different technologies into one platform. In the top layer, which is the user class, n
đang được dịch, vui lòng đợi..
