Item: Oracle Database
Rating: 9
Author: yixiang Shan

Use Cases and Deployment Scope

Oracle Database, has been used as the Enterprise Data Storage solution for many years in our company. Even today, when a lot of new NoSQL database products pop up, more than 80% of corporate and regional data is still stored in Oracle Database. Oracle Database serves us well in most of the business use cases, performance wise and stability wise, in different domains like operation, clearance, reporting and post ops analysis.

Pros and Cons

We have databases in two data centers geographically located in different places in our region. The Oracle 12c Data Guard feature makes sure the data instance is synchronized in a near real-time manner from the primary instance to the standby instance, across the data centers to provide data center level failure resilience.
Database instance's high availability is guaranteed by the RAC feature, which also helps us scale up resources like servers, CPU and memory.
Performance tuning toolset provided by Oracle Database is quite handy.
The security feature is extremely important for our company, and we are quite happy about what is offered by Oracle Database.

Oracle DB doesn't work very well for dealing with a large amount of time series based data rows performance wise.
The traditional primary/standby mode of data guard architecture, causing half of the resources (Stand-By side) to be idle most of the time and become a wasted investment. In the future, the near real-time data sync should be made available on the standby side and should allow the read-only operation, so at least business can use the resources for data analytic purposes. That will be great
When traditional relational database architecture is challenged by the NoSQL Technics, in most of the case, the business or the end user is not looking for some "Oracle DB" replacement solution, instead, they are seeking a "hybrid" solution. As they get used to the Oracle environment and they like the ACID operation, what they are looking for the flexibility features that are provided by pure NoSQL products (like variable columns for different rows in the same table etc). Oracle has some improvements like security file on BLOB, but if Oracle can make more improvements aiming in this direction, that can bring more added value and retain a lot of old users
For some use cases, we noticed that database performance deteriorating if we have relatively high active connection sessions

Return on Investment

In the express industry, Oracle is still considered as the best and most reliable data storage solution
The only negative point is also well known, the license cost of Oracle, making a lot of small or medium companies opt for some cheaper and open source solutions

Alternatives Considered

MySQL, PostgreSQL, Cassandra and Neo4j

Oracle completed solution and support is of cause better than open source solutions, we also choose other NoSQL database and Graph database, to address some very particular business use cases

Other Software Used

TensorFlow, Caffe Deep Learning Framework, Hadoop

Likelihood to Recommend

Oracle Database 12c is very suitable for the Enterprise to deal with well defined structured data, requiring a good balance between the data set and the query performance. It is also well suited for during data insertion time when it is hard to predict how the query will be made (index later strategy). If the business requires highly reliable transaction-based ACID operations and high-security standards, Oracle is one of the best options. While Oracle is not suitable for applying data analytic tasks to extremely large amounts of data (which is not Oracle's issue, it is SQL's limitation), a distributed solution like Hadoop or Cassandra should be considered.

Oracle Database 12c: one of the best enterprise data storage solutions

Overall Satisfaction with Oracle Database 12c