What is most valuable?
It provides very fast query performance after good designs of projections.
It's easy to implement for 24/7 data load and usage because you don’t have to worry about “load time slots” since you can load data into reporting tables at all times without worrying about their query load.
It just keeps up and running all the time.
How has it helped my organization?
We have been able to move from nightly batch loads to continuous data flow and usage. This hasn’t happened just because of Vertica, we have renewed our data platform pretty thoroughly, but definitely Vertica is one major part of our new data platform.
What needs improvement?
We are running our data transformations as an ELT process inside Vertica; we have data at least on the landing area, temporary staging area, and final data model. Data transformations require lots of deletes and updates (which are actually delete/insets in Vertica). Delete in Vertica doesn’t actually delete data from tables, it just marks them as deleted. For us to keep the performance up, purge procedures are needed and a good delete strategy needs to be designed and implemented. This can be time consuming and is a hard task to complete, so more ‘out-of-the-box’ delete strategies would be a nice improvement.
For how long have I used the solution?
We've been using it since January 2015.
What was my experience with deployment of the solution?
We haven't had any issues with the deployment.
What do I think about the stability of the solution?
Stability is good, however the database crashed once because a query ran against a large XML data element.
What do I think about the scalability of the solution?
We haven’t yet scaled out our system. So far performance has been good (taking into consideration that delete strategy mentioned in the Areas for Improvement question).
How are customer service and technical support?
We haven’t needed tech support too much. So far so good.
Which solution did I use previously and why did I switch?
We used Oracle for our DWH. When selecting a new database, we evaluated -- based both on written documentation and hands-on experimenting -- quite a lot of databases, such as Exadata, Teradata, and IBM Netezza. We selected HP Vertica as it runs on bulk hardware since it has “open interfaces”. It performed really well during hands-on experimenting and its “theories in practice” is good. Performance is excellent, development is easy (however, you need to re-think some things that you may have gotten used to when using other SQL databases), and its license model is simple.
How was the initial setup?
It seemed to be very straightforward. However, we had an experienced consult to do the setup.
What about the implementation team?
We had a joint team consisting of both an in-house team and external consultants. It’s very important to build up the internal knowledge by participating in actual project work.
What was our ROI?
We have ran so little time in production that we don’t yet have a decent ROI or other calculations done.
What's my experience with pricing, setup cost, and licensing?
The license model of HP Vertica is simple and transparent.
What other advice do I have?
Just go for it and try it out; you can download the free Community edition from the HP Vertica website.