The biggest lesson that I have learned from using this solution is that a lot of evaluation should be done before starting. Also, we needed to put a lot of effort into understanding the different functions that the product offers. This allows you to best leverage the capability of the product.

I would rate this solution a seven out of ten.

The first thing is, do a full PoC where you actually plug in dummy data. Start with millions of records and try out the options which you have already in your environment and see if you really get a benefit in terms of performance. Because this product is mainly to improve the performance.

There are two things. First, it improves the performance and then it's easy to maintain. When I say easy, you would not need a DBA at all. Anybody who has really good knowledge in SQL should be able to maintain this product. That cost, of a DBA, you will completely be eliminating.

The first part of this product is basically performance. If you are moving on from an ODBC database to this, then you would need to do a round of testing on data because you are actually moving from a different technology, that is, to an appliance technology. There are multiple things which can stop you, things you might not be able to do; operations, like booking of a ticket or reloading a feed. We tried to do those things here, in Oracle or MySQL, but when it comes to reporting, this is the best solution we have so far. By that I mean, cheapest and the best.

Then, try to figure out if you really require a code change, because there is a minimal time for a code change; but if it's a request then you should catch it when you're doing a PoC.

Take out any complex situation, give it to the vendor, IBM. Tell them to do a small PoC on the dummy data and let them come up with the performance, then you can compare with your feeds.

Overall, it has good connectivity with almost all the niche technologies we have on the market right now. It can scale up to any technology, any content. That is another benefit of having this.

IBM is a good company which always tries to be competitive with other technologies on the market. If we are stuck, we are trying to purchase something for our content and we are stuck somewhere, something is missing, we approach them and these guys will give us the solution in terms of a patch or an API or something like that, which will help us connect to the newest technologies.

I would have to understand the specifics of the organization, the data size, the tasks, etc. to make any recommendations.

We were using the on-premises deployment model.

I haven't been following the solution for the last year or year and a half, so I don't have too much to say about it.

I'd rate it seven out of ten.

I don’t want to be biased but it all depends on the business requirement because there are many models for it, which range from 7.8 TB to 1.2 PB (uncompressed). If the business has done a feasibility check on Netezza for their scalability plus the growth and they found it to be within their estimates then I would suggest that they have taken off from the right runway.

I don't think that it is currently available on the cloud, except for IBM Bluemix. They will come up with DB2 Analytics. IBM will come up with a new name.

Moving the data from one box to another box or moving it to the cloud or any other storage, the connectors are there but the efficiency in the movement or the data movement features is similar to USB.

Netezza has already changed its name, so they are no longer selling this solution under this brand. They are selling Neteeza Mako and I would recommend it, although I am still looking for an analytics solution.

Netezza has already launched the new DB2 analytics.

You want to think about whether or not you want to purchase it or consider other solutions such as Snowflake, and Redshift.

When we deployed this solution, the FPGA concept was not at all clear until we started using FPGA with SQL performance-tuning that combines the hardware with the software. It's tightly coupled.

I was not aware of the Sqoop connectors to Hadoop that are used to transfer data from databases to Hadoop and vice-versa. We already had Neteeza, so we only had to download Sqoop. There are several SQL analytics. Not everyone is aware that these functions are already built and that there are separate analytics and SQL analytics function libraries available to download for free.

I would rate this solution a seven out of ten.

For the last two years, we've been using the cloud deployment model. We're an IBM partner.

We're using various types of databases including DB2 and Microsoft SQL.

I'd rate the solution ten out of ten. We had no problem integrating this solution with our legacy platform. 

Make sure to check the capability regarding in-DB processing with the application you will be using with this appliance. I am using SAS and it's a dis-appointment due to limited in-DB processing, including data connection pool, that was never resolved by both SAS and IBM Netezza.

You should know how it works well so you can use its power. So, start with a partner or vendor team.

We use the private cloud deployment model.

I would rate the solution six out of ten.