Amazon Elastic MapReduce is Worth the Effort if You Really Need Big Data Management, Just Make Sure You Really Need It
February 04, 2019
Amazon Elastic MapReduce is Worth the Effort if You Really Need Big Data Management, Just Make Sure You Really Need It
Score 8 out of 10
Vetted Review
Verified User
Overall Satisfaction with Amazon Elastic MapReduce
Amazon Elastic MapReduce is used by my department to produce big data analytics for certain clients. The software address data mining and predictive analytics for data sets that take a long time to process. The software is not used for econometric or other analytical evaluation because the size of the data sets does not lend themselves to such analysis. The software is used almost exclusively for data mining and simple reporting for large data cases.
- Amazon Elastic MapReduce works well for managing analyses that use multiple tools, such as Hadoop and Spark. If it were not for the fact that we use multiple tools, there would be less need for MapReduce.
- MapReduce is always on. I've never had a problem getting data analyses to run on the system. It's simple to set up data mining projects.
- Amazon Elastic MapReduce has no problems dealing with very large data sets. It processes them just fine. With that said, the outputs don't come instantaneously. It takes time.
- The analytical processes generally run quicker with the standalone tools of Hadoop, Spark, and others. If you only use one big data tool and don't really need things simplified, then Elastic MapReduce is more of an overhead tool that doesn't add much value.
- The analytical capabilities of Elastic MapReduce are nowhere near as complex or broad as non-big data tools. I would suggest not using the tool unless your data really is big data.
- The machine learning capabilities of Elastic MapReduce (using the big data tools of Hadoop/Spark) are good but are not as easy to use as other machine learning tools.
- Amazon Elastic MapReduce has had a positive ROI in the sense that it saved time managing big data projects where analysts were using different big data tools. Essentially, an increase in employee productivity.
- Elastic MapReduce is not worth it in cases where you're just trying things out. You'll likely lose money unless you're sure that using MapReduce is a good idea.
- Elastic MapReduce takes some time learning, although not too much. If the employee is less well-versed in big data analytics, the software is a high hill to climb that eats up employee time.
- Hadoop, Apache Spark, Apache Spark MLib, Apache Spark Streaming, Apache Web Server, Google Cloud AI, Google Cloud CDN, Google Cloud Dataflow, Google Cloud Datastore, Google Cloud Pub/Sub, Google Cloud SQL, Google Cloud Spanner, Google Data Studio, Google Correlate, Google Marketing Platform (formerly DoubleClick), Google Pay (formerly Google Wallet), Hortonworks Data Platform, Cloudera Data Science Workbench, Cloudera Enterprise, Cloudera Manager, Datameer, Pi Datametrics, FICO Model Builder, IBM SPSS Modeler, Microsoft Dynamics SL, Microsoft Exchange, 6sense and Sybase Aleri Streaming Platform
Perhaps the biggest advantage Amazon Elastic MapReduce has over competing big data management software is the user base. Elastic MapReduce, compliments of its connection with Amazon, has a large user base to whom questions about functionality can be addressed. The software also has a very nice user interface. Additionally, Elastic MapReduce runs fairly quickly and the results are generally easier to manipulate. With this said, Elastic MapReduce is definitely not the easiest nor quickest tool for big data analytics.