1.黑斑羚：
Impala是在Hadoop上运行的查询引擎。它是一个开源软件，可以大规模并行处理SQL查询引擎。它支持内存中的数据处理。它开创了Parquet文件格式的使用，Parquet文件格式是一种列式存储布局，已针对数据仓库场景中典型的大规模查询进行了优化。它提供高性能，低延迟的SQL查询，还提供对以Hadoop文件格式存储的数据的交互式查询处理。
2.甲骨文：
Oracle是由Oracle Corporation开发的商业软件。 Oracle广泛使用RDBMS。 Oracle允许快速安全地存储和检索数据。它用于运行在线事务处理和数据仓库。 Oracle在Mac OS，UNIX，Windows和Linux等最主要的操作系统上运行。 Oracle具有网络堆栈，允许来自不同平台的应用程序轻松与Oracle通信。
Impala和Oracle之间的区别：
| S.NO. | Impala | Oracle |
|---|---|---|
| 1. | It was developed by Cloudera. | It was developed by Oracle Corporation. |
| 2. | It was launched in 2013. | It was launched in 1980. |
| 3. | It is an open source software. | It is commercial. |
| 4. | Impala is implemented in C++ language. | Oracle is implemented in C and C++ languages. |
| 5. | It does not support XML. | XML support is available. |
| 6. | JDBC and ODBC are the APIs and access methods used in Impala. | JDBC, ODBC, ODP.NET and Oracle Call Interface (OCI) are the APIs and access methods used in Oracle. |
| 7. | It uses Sharding partitioning methods for storing different data on different nodes. | It uses Horizontal partitioning with the Optional Oracle Partitioning methods. |
| 8. | No concept of referential integrity (Foreign key). | Referential integrity is used in Oracle. |
| 9. | No transaction concepts. | ACID properties of transaction is used. |
| 10. | There is no option to define some or all structures to be held in-memory only. | It provide option to define some or all structures to be held in-memory only. |