Vertica Review

Vertica allows for thousands of users to run an analysis at the same time. Great aggressive compression.

At the tech company I work , we were looking for new ways to allow end users (a couple of thousand external users)  to crunch through their detailed data in real time as well as enabling internal users and data analysts to gain the information they needed to run and optimize their business processes. 

Unfortunately our current system had  become slower and slower over time due to the tremendous increase in data to be managed so a new approach had to be taken to accomplish this goal.  Our existing data warehouse/data management infrastructure just could not handle big data.

We evaluated a variety of different solutions such as Amazon Redshift, Infobright and Microsoft. Vertica won out above all these other solutions. Our dataset is several hundred million rows and our avg. response time goal was less than 5 secs. We are building our environment for the future so another requirement was to be able to scale horizontally. 

Redshift came close in response time but failed in concurrency, meaning multiple users running an analysis at the same time. Infobright came close in response time and concurrency but didn’t provide sufficient scalability. Vertica checked all boxes at a very competitive price-point.

We found that the extreme speed, performance and flexibility is superior to all the other solutions out there. The massive scalability on industry-standard hardware, standard SQL interface and database designer and administration tools are excellent features of Vertica. I also really value the simplicity, concurrency for hundreds or thousands of users, and aggressive compression.

This new environment allowed us to implement applications such as clickstream and predictive analysis which have added tremendous value for us. Currently there is about 500 GB – 1 TB of data that I am managing and I have found that Vertica is able to be integrated very well with a variety of Business Intelligence (BI), visualization, and ETL tools in their environment. I use Hadoop, Tableau and Birst and using all these solutions with Vertica has been overall quite smooth.

Our query performance has increased by 500 – 1,000% through improvements in response time and I am now able to compress our data by more than 50%. The simultaneous loading and querying and aggressive compression has helped us become more efficient and productive. Furthermore the high availability without hardware redundancy, optimizer and execution engine, and high availability for analytics systems has saved us both time and money. 

**Disclosure: IT Central Station has made contact with the reviewer to validate that the person is a real user. The information in the posting is based upon a vendor-supplied case study, but the reviewer has confirmed the content's accuracy.
More Vertica reviews from users
...who work at a Computer Software Company
...who compared it with Teradata
Learn what your peers think about Vertica. Get advice and tips from experienced pros sharing their opinions. Updated: April 2021.
479,323 professionals have used our research since 2012.
Add a Comment

author avatarit_user3876 (Database Manager at a tech services company with 501-1,000 employees)

Hi Robert. Does Vertica provide any configuration options for limiting database size and resource usage?

author avatarit_user107709 (Architect at a tech services company with 51-200 employees)

Yes it does. You can create performance profiles limiting users to consume only certain amount of memory and cpu. Secondly licensing is done via database size.

author avatarit_user110223 (Database Expert at a tech services company)

I think @imran_sh is asking about space usage y object.Like the tablespace concept from Oracle as an example.

Yes, Vertica has a good Resource Manager where you can create custom resource pools for you different user profiles. The negative part ! - no dynamic resource pools yet.

author avatarit_user124989 (User)

It seems you were mainly focused on how Vertica is good and did not run a benchmark, otherwise it could be nice if you could publish loading and query performance between all above DB's. 500GB - 1TB is not a lot of data .