Apache Hive

Apache Hive is an open source software which is very powerful for helps you manage large datasets and enable efficient querying in distributed storage developed by Facebook and used by large companies like Netflix as well. Its infrastructure is built on top of Hadoop which already allows for large dataset processing across vast networks of machines with massive storage capabilities with each in the network with local computing and local storage capabilities at the same time. Which just kind of gives you an idea of how large we are talking about in terms of the framework Apache Hive works with including Hadoop integration. The HiveQL language is quite similar to SQL query language though it is still different from standard SQL. It can also be extended quite a lot with aggregations, custom scalar functions and table functions while defining your own Hive SerDe formatted data. The ability to extended Hive with scalar functions is good. Hive is designed for batch jobs using super sized sets of web log like data to be appended only as it doesn't have row level updating or real time querying. Being open source you have access to a very active community and can ask regular users and find experts who will help you along as well. There is a lot documentation around though it may seem somewhat scattered as open source docs can tend to do. But there is a lot of great help spread out on forum threads as well. Stack Exchange is a good site to look for help as well. As an open source software you can never complain about the price tag. Free in every sense of the word is a pretty cool concept.

The code base for Hive is really large and it certainly would take an investment to support and maintain ongoing with so many configurations possible. There is most certainly quite a learning curve with Hive if you've used other similar software like Pig and SQL this will help but it is still it's own format and requires time to fully utilize. As it is open source software you may have trouble getting the level of customer support you need from the company itself and this will require you to reach out externally to devs who can help and this is another cost to consider.

This data warehouse infrastructure was built for massive scale on thousands of machines and does a good job of summarizing, analyzing and querying data in distributed file storage. It provides a way for data extraction, transformation, and load while it can be run on both HDFS and Hadoop. You can use Hive for business intelligence as well if you reuse your SQL. And the syntax is quite similar to SQL so you can pick it up if you are already familiar with it relatively quickly. Hive seems to be quite popular for analysis of large datasets to help with forecasting and projections, debugging issues, business intelligence and so on and works well with Pig for this purpose.