What is Scalability in Cloud Computing

What is Scalability in Cloud Computing

If you use Hive on Hadoop, learning about Hive queries—and getting familiar with a Hive query example or two—is key to achieving effective cluster management. Hive queries consume time and resources, so achieving efficiency through Hive query tuning is a must. In this article, you’ll learn what Hive queries are, how they can affect your clusters (both positively and negatively), useful Hive query approaches, and what a good Hive query example looks like. Let’s get started.

1. What Are Hive Queries?

Hive queries are specific information requests from your Hadoop database. These information requests are performed by Apache Hive, an open-source data warehousing platform developed on top of Hadoop. Facebook created Hive to perform data analysis, distributed processing, and reduce work in terms of writing the Java MapReduce platform.

A Hive query utilizes a set of predefined codes native to your database’s language. The database then receives the instruction, and once it understands that instruction, gathers and releases the information requested.

Hive was made for efficiency, and that is why its queries need to be perfectly tuned and well-written. You can also set dependencies to enable the auto-scheduling of queries. This will guarantee that as soon as one action is finished, the next action starts immediately.

Improving the RAM capacity and CPU power of your system results in faster Hive response times, compared to increasing your network bandwidth.

2. The Effects of Poorly-Tuned Hive Queries

Poorly-tuned queries can cause major setbacks for an organization. The biggest of these would be missed SLAs (service level agreements).

These agreements signify the level of service that enterprises and their clients agree upon, and they include performance guarantees, data security, uptimes, and customer service standards. So if inefficient queries lead to missed SLAs, the result can be penalties, refunds, or, in some cases, termination of the contract.

Poorly tuned Hive queries also consume resources. These can affect your Hadoop clusters on two fronts. One: Poorly-tuned queries can use up resources intended for other users or functionalities in your cluster. This results in reduced performance and slower response times. Two: The resources used incur costs. Wasted resources due to poorly-tuned Hive queries can add up on your AWS bill and give you a major headache.

Some other effects of inefficient queries can be disrupted cluster performance, slowing down the database, and downtimes.

Because of the numerous negative effects inefficient queries can create, it’s crucial to optimize your queries. And while you can use manual approaches like partitioning, bucketing, and compression, leveraging analytics stack performance tools like Pepperdata Query Spotlight will make the job a lot easier.

3. Hive Query Tuning Approaches

There are some extremely handy Hive query tuning approaches, depending on whether you’re optimizing for time or resource usage:


Proper Hive tuning allows you to manipulate as little data as possible. One way to do this is through partitioning, where you assign “keys” to subdirectories where your data is segregated. When your query asks for information, you can target the specific subset where your data is, which saves you time from scanning data you don’t need.


Bucketing is similar to partitioning, albeit it helps to improve join performance by scanning less data.


This helps minimize the amount of the