Spark SQL Reviews & Product Details

Specifically talking the few best points I like about Spark SQL is as follows:

- It is the best choice for big data analytics in collaboration with Hadoop.

- It provides fast access to data in SQL workloads.

- In Spark SQL, many types of data processing can be used together.

- It is easy to pull in multiple data sources - from Spark RDD to external databases.

- Spark SQL supports Map-reduce, SQL queries, Streaming data, Machine learning (ML), and Graph algorithms. Review collected by and hosted on G2.com.

My major dislike is Spark SQL's limitations, including Latency issues, minor files issues, and no real-time data processing. Apache has already resolved some with an alternative solution by Apache Apex. However, These issues need to be determined at Spark SQL as an alternative is okay, but some features that Spark SQL offers aren't available with Apex. Review collected by and hosted on G2.com.

1. Query big data with SQL syntax.

2. Perform aggregations. Review collected by and hosted on G2.com.

Setting up standalone spark cluster is difficult. Review collected by and hosted on G2.com.

One feature that I like very much as a software inclined person is that Spark is open source. This allows me to dig deeper into its value than other software that is out there that perform the same tasks. It is also extremely robust in functionality as software. Even more than that this framework and SQL version is extremely faster than its competitors because of the way that it processes data. Spark uses this idea called cluster mode and within that, it uses distributed processing and it allows for the computation in the frameworks to work entirely internally. Another great feature is the ability to source data from multiple different sources. The main use that I can see within this framework is machine learning and AI when you are using large amounts of data from multiple sources. It allows the best combinability and processing ability. Review collected by and hosted on G2.com.

One thing that I am not fond of with this software is that there is no dataset API support in the python version of this software. The reason this is an issue is that most machine learning and AI programs that I write are in the Python language, so without that support, it is a heavy downfall for me personally. It still works, but without that feature, it is a small letdown. In terms of other programming languages, it works great. Another potential issue is that there is an inability to support multiple concurrent users at a time which can provide potential roadblocks on big projects. Review collected by and hosted on G2.com.

Ability to handle big data and display the result in the least time. Review collected by and hosted on G2.com.

Noting. Faced no issue while using the same. Review collected by and hosted on G2.com.

Some features that I like most about spark SQL are-

Spark SQL is so smoothly integrated with big data tools that it makes it a go-to product.

It can process a large amount of data in seconds, that the efficiency.

Its query syntax are like regular SQL only which reduces efforts to learn a new language. Review collected by and hosted on G2.com.

Spark sql needs huge amount of RAM for processing data that maked too mamuch memory consumption.

Also we can't create table in spark for union type. Review collected by and hosted on G2.com.

Feasability smd easy understanding of the queries built inside sparksql Review collected by and hosted on G2.com.

Non friendly ui is very unsatisfying, it has to be improved Review collected by and hosted on G2.com.

Spark sql is a spark module for structured data processing Review collected by and hosted on G2.com.

No file management system,expensive,window criteria Review collected by and hosted on G2.com.

Its support is to access different data sources like Hive, Avro, Parquet, ORC, JSON, and JDBC. Review collected by and hosted on G2.com.

It Doesn't support real-time processing. Review collected by and hosted on G2.com.

1. We can write most of the SQL scripts to read the massive data

2. All kinds of SQL features are supported

3. The best feature I like is to use the SQL in python language and store the extract the data in python data frames

4. SQL views are very comfortable way to create access the data from data lakes

5. Since Spark SQL uses the Spark clusters it is very fast to run the data extraction related queries

6. Notebooks are more easy and comfortable way to write the code Review collected by and hosted on G2.com.

Only problem is create tables or views on top of the JSON data from data lake. And to setup the external meta store instead of upgrading the current metastore Review collected by and hosted on G2.com.

It enables to run Hadoop-Hive queries 10x times faster than MR. It allows you to query data frames in python, scala, or java using queries in SQL. Review collected by and hosted on G2.com.

There is nothing I don't like about Spark SQL. It is best to query big data in a fast approach. Review collected by and hosted on G2.com.