The fact that it is a columnar database is valuable. Columnar storage has its own benefit with a large amount of data. It's superior to most traditional relational DB when dealing with a large amount of data. We believe that Vertica is one of the best players in this realm.
Improvements to My Organization
Large-volume queries are executed in a relatively short amount of time, so that we could develop reports that consume data in Vertica.
Room for Improvement
Speed: It's already doing what it is supposed to do in terms of speed but still, as a user, I hope it gets even faster.
Specific to our company, we do store the data both in AWS S3 and Vertica. For some batch jobs, we decided to create a Spark job rather than Vertica operations for speed and/or scalability concerns. Maybe this is just due to the computation efficiency between SQL operations vs. a programmatic approach. Even with some optimization (adding projections for merge joins and grouped by pipelined), it's still taking a longer time than a Spark job in some cases.
Use of Solution
I have personally used it for about 2.5 years.
I have not recently encountered any stability issues; we have good health checks/monitoring around Vertica now.
I have not encountered any scalability issues; I think it's scalable.
Customer Service and Technical Support
N/A; don't have much experience on this.
We do have some pipelines accessing raw data directly and process it as a batch Spark job. Why? I guess it's because the type of operations we do can be done easily in code vs. SQL.
I would recommend using Vertica for those people/teams having large denormalized fact tables that need to be processed efficiently. I worked around optimizing the query performance dealing with projections, merge joins and groupby pipelines. It paid off at the end.
Disclosure: I am a real user, and this review is based on my own experience and opinions.
Oct 18 2016