Introduction to HiveQL
HiveQL, commonly known as HQL, is a SQL- like query language for processing large chunks of data from the Hadoop environment. Originally developed by Facebook around 2007, Hive has later become an open-source Apache project. Hive queries with syntax similar to SQL have become popular among data analysts nowadays. HiveQL using HDFS for storage and MapReduce for processing helps in deriving insights for big data users with minimum coding experience. Various other distributed computation engines like Apache Tez and Apache Spark can be used with HiveQL for faster processing of large-scale data analytics. The various ways to use HiveQL are – CLI (Command Line Interface), commercial products like Karmasphere, Cloudera’s open source Hue and Qubole.
Working of HiveQL
Hive provides a SQL like dialect for data manipulation, eliminating the need for us to write low-level MapReduce jobs to fetch data (through Mapper) and aggregate final results (through Reducer Modules).
Executing Hive Query
Steps involved in executing a Hive query are:
- Hive interface (through CLI or Web UI) sends the query to Driver (which is JDBC, ODBC or Thrift Server) to compile, optimize and execute. It checks the query through a compiler for syntax and execution plan.
- Compiler sends request to Metastore and receives Metadata as a response.
- The compiler communicates the execution plan back to the driver, which further sends it to the Execution engine.
- The execution engine interacts with the Job Tracker, which is the Name Node to get the job done.
- In parallel, execution engine also executes Metadata operation.
- Name node gets the job done by Task tracker or Data nodes.
- Results that are shared with the execution engine, which then are displayed to the interface through the driver.
Optimization of Queries
Tuning HiveQL for better optimization of queries. Using the below-set commands, we can override the default configurations and enable faster query execution.
1. SET hive.execution.engine = tez
By default, the execution engine is set as Mapreduce, but we can explicitly set it to tez (for Hadoop 2 only) or Spark (for Hive 1.1. 0 onward).
2. SET hive.mapred.mode = unstrict
This is for dynamic partitioning which is helpful for loading large data sets. However, static partitioning is set as default, which happens in “strict” mode.
3. SET hive.vectorized.execution = true
set hive.vectorized.execution.enabled = true
Vectorized query execution allows operation like –aggregates, filters or joins to happen in batches of 1024 rows instead of single row at a time.
4.5 (2,454 ratings)
View Course
4. SET hive.auto.convert.join = true
While joining a large data set with a very small data set, map joins are more efficient and can be set using the above command.
5. SET hive.exec.parallel = true
MapReduce jobs are executed in parallel in Hadoop. Sometimes, if queries are not dependent on one another, parallel execution can be favored leading to better memory management.
6. SET hive.exec.compress.output=true
This enables the final output to be stored in HDFS in a compressed format.
7. SET hive.exec.compress.output = true
This enables the final output to be stored in HDFS in a compressed format.
Features of HiveQL
- Being a high-level language, Hive queries are implicitly converted to map-reduce jobs or complex DAGs (directed acyclic graphs). Using the ‘Explain’ keyword before the query, we can get the query plan.
- Faster query execution using Metadata storage in an RDMS format and replicates data, making retrieval easy in case of loss.
- Bitmap indexing is done to accelerate queries.
- Enhances performance by allowing partitioning of data.
- Hive can process different types of compressed files, thus saving disk space.
- To manipulate strings, integers or dates, HiveQL supports extending the user-defined functions (UDF), to solve problems not supported by built-in UDFs.
- It provides a range of additional APIs, to build a customized query engine.
- Different file formats are supported like Textfile, Sequencefile, ORC(Optimised Row Columnar), RCFile, Avro and Parquet. ORC file format is most suitable for improving query performance as it stores data in the most optimized way, leading to faster query execution.
- It is an efficient data analytics and ETL tool for large datasets 10. Easy to write queries as it is similar to SQL. DDL (Data definition language) commands in hive are used to specify and change the structure of the database or tables in hive. These commands are drop, create, truncate, alter, show or describe.
Limitations
- Hive queries have higher latency as Hadoop is a batch-oriented system.
- Nested or sub-queries are not supported.
- Update or delete or insert operation cannot be performed at a record level.
- Real-time data processing or querying is not offered through the Hive Scope of HQL.
With petabytes of data, ranging from billions to trillions of record, HiveQL has a large scope for big data professionals.
Scope of HiveQL
Below are the ways in which the scope of HiveQL widens and better serves the purpose of analyzing humungous data generated by users every day.
Security: Along with processing large data, Hive provides data security. This task is complex for the distributed system, as multiple components are needed to communicate with each other. Kerberos authorization support allows authentication between client and server.
Locking: Traditionally, Hive lacks locking on rows, columns or queries. Hive can leverage Apache Zookeeper for locking support.
Workflow Management: Apache Oozie is a workflow scheduler to automate various HiveQL queries to execute sequentially or in parallel.
Visualization: Zeppelin notebook is a web-based notebook, which enables interactive data analytics. It supports Hive and Spark for data visualization and collaboration.
Conclusion
HiveQL is widely used across organizations to solve complex use cases. Keeping in mind the features and limitations offered by the language, Hive query language is used in telecommunications, healthcare, retail, banking, and financial services and even NASA’s Test Propulsion Laboratory’s Climate evaluation system. Ease of writing SQL like queries and commands accounts for wider acceptance. The growing job opportunity in this field lures fresher and professionals from different sectors to gain hands-on experience and knowledge about the field.
Recommended Articles
This is a guide to HiveQL. Here we discuss the Introduction to HiveQL, Optimizing Queries and its Limitations and Features. You can also go through our related articles to learn more –
