Introducing Presto 0.193 in QDS

Start Free Trial
June 1, 2018 by Updated May 19th, 2022

At Qubole, it has been our constant endeavor to ensure that our customers benefit from the latest improvements and features in open source releases while continuing to use value additions from Qubole. Continuing that tradition, we are pleased to announce that Presto 0.193 (Beta) is now available on Qubole Data Service. With Qubole Presto 0.193, our customers can continue leveraging the following Qubole value additions from previous Qubole Presto releases:

1. Autoscaling

In a recent blog, we demonstrated how autoscaling plays an important role in major cost savings for many Qubole customers. Qubole Presto 0.193 will support autoscaling, just like our previously supported presto versions.

2. UDF support

We open-sourced Presto UDF project which has many commonly used user-defined functions for Presto. This would be available for Qubole Presto 0.193 too. More details about UDFs are available here.

3. Rubix

Rubix is a Next-generation disk cache for cloud storage systems. We saw improvements as big as 8x, with improvements being directly proportional to the amount of data being read. Rubix integration will be available with Qubole Presto 0.193.

4. Join Optimizations

Join reordering and Dynamic Filtering are two join optimizations introduced in Qubole Presto, which can improve query times by up to 14x. More details can be found on our blog.

5. Faster Ruby Client

Ruby client has been integrated with Qubole Presto which would be available for Qubole Presto 0.193 offering too. We have seen good improvements in query times due to Ruby’s client and more details can be found here.

6. Presto Kinesis Connector

Qubole has developed the Presto Kinesis connector. The connector allows the use of Kinesis streams as tables in Presto, such that each data blob is presented as a row in presto. More details can be found on our Github page.

In addition to the above Qubole value additions, Qubole Presto 0.193 comes with open-source features and improvements such as:

  1. Resource Groups
  2. Spill to Disk
  3. Stability Improvements
  4. Geospatial Functions (This feature has also been backported to Qubole Presto 0.180 due to active customer demands.)
  5. Redshift connector

You can launch a cluster with Presto 0.193 (beta) just like any other supported Presto version. Please see below for the screenshot:

Please try it out and drop your feedback with the Qubole representative you are in touch with or at [email protected].

 

Start Free Trial
  • Recent Posts

  • Categories

  • Events

    Data2030 Summit 2022 – APAC Edition – Data Strategies For Data And AI-Driven Organisations

    May. 25, 2022 | Global

    Faster Interactive and Adhoc Queries with Qubole Presto

    May. 26, 2022 | Online

    Stop the Costs of Big Data Projects Spiraling Out of Control

    Jun. 2, 2022 | Online

    Federated Queries are the Future of Analytics

    Jun. 9, 2022 | Online

    Workload-Aware Autoscaling Reduces Data Lake Costs

    Jun. 16, 2022 | Online
  • Read Success with Predictive Analytics—Getting the Basics Right