We just raised a $30M Series A: Read our story

When evaluating Data Warehouse solutions, what aspect do you think is the most important to look for?


Let the community know what you think. Share your opinions now!

ITCS user
1212 Answers

author avatar

It depends on the company requirement. Performance of database when loading and retrieving data is very important for any data warehouse.

author avatar

I would look at it from how the solution addresses two types of requirements - call it Functional and Non-functional or Business and Technical requirements.
This classification becomes critical as the datawarehouse solutions have now evolved from mere high-power databases with enhanced storage and performance optimization features to a key Information management solution in the enterprise.

From Business perspective, look at:
- Analytical capability it can support (likes of SAP HANA have all the analytical layers built within the datawarehouse not needing a high-end analytical tool!)
- Variety of data to support (structured and unstructured - how they can co-exist with the big data solutions)
- Capability to support or integrate with other data management solutions like Master Data Management, Data Governance, Data Quality solutions (isolated DWH solution does not show any business value and will die a slow death)
- Ability to provide or support accelerators such as out-of-the-box industry data models and other Agile development methods (avoids DWH projects becoming multi-year projects delivering no ROI)

From Technical perspective, look at:
- Total cost of ownership - Cloud-based vs appliances vs on-prem database platform
- Scalability and Performance features - in-memory architecture, row-based vs columnar, parallel processing(MPP) arch etc
- Storage optimization - data compression capability, caching etc
- Security (critical for regulated industries) such as support for masking, advance audit techniques and access control

The key point I want to bring up again is that it is just not the technical perspective that should drive the selection.

author avatar
Real User

Ease of use in finding and retrieving data. Having the data stored without providing the meta data for the user is of no value. it needs an interface with access to data that provides more than key word search.

author avatar

A very open-ended question.
When evaluating a DW initiative, the most important thing is to start off with the right questions, many of which have already been touched-on directly or indirectly by others here.

Know your data and what you want to do with it; how much do you already have, and for how long do you want to keep it in the DW?
How many people will be querying it at the same time?
What kind of response time do they expect?
Which DW platforms are compatible with your existing IT staff skillset? What skillsets would you have to hire?
Is the data ready to be loaded into the DW? It seldom is - will additional staff be needed for this?

author avatar

About my experience, and seeing how the TICs are moving today, I think that there are two principals factors to consider when one wants to invest in a DW platform. These are: Innovation and support...The first assure you about a continuous modern platform over the years, and the second assure you that this platform runs at the level you want.

author avatar

It actually boils down to the amount of Sequential IO that can be pushed and massaged. Each physical core of a data warehousing host can consume about 1 gbps of data... In order to keep all the cores working on the data it has to get moved from storage into ram... you need to measure the Maximum Consumption rates and the Maximum Throughput rates for Sequential IO and determine whether there are any bottlenecks to the performance of queries acting against the sequential data.

author avatar
Top 20Real User

IMHO, it is the wrong approach to consider one single aspect as the driver for comparing DW solutions from vendors.

Over the years, many aspects have become important for a DW solution:

ETL tools included in the solution, in particular, tools must include required transformations for Dimensions and Fact Tables, like Slowly Changing Dimension, just to name one

Data Quality tools included in the solution

Master Data Services tools included in the solution

Reporting tools included in the solution
Level of integration between ETL tools, Data Quality tools and Master Data Services tools included in the solution

Proper extensibility support in ETL tools included in the solution

Proper table partitioning (data or horizontal partitioning) support included in the solution

Proper columnstore indexes support included in the solution

These are the important aspects that a competitive DW solution should include.

author avatar
Top 5PopularReal User

It 100% depends on your needs in terms of data volume, availability, # of potential concurrent users, # of data sources, complexity of data sources. I'm assuming by platform you mean tools/software/hardware? Are you thinking BI as well or just backend?

Cloud-based solutions should be part of the consideration but that again is based on your organization's needs and abilities - data center capacity and support. Cost always matters no matter where you work and we all have budgets to work with so saying scalability is most important or something else doesn't make sense if you don't need scalability or can't afford it.

In other words, if your need is just solving reporting problems for a single source system (not really a DW), your source is only say 100 GB and you have a dozen total users, then buying 10 node Teradata MPP means you have lost your marbles. So scalability in that situation is irrelevant because almost anything will satisfy your needs.

On the other hand if you may have a petabyte of total data and terabyte DW database then get out your wallet and buy top of the line...Informatica or Datastage are probably best bets for ETL, and Teradata or maybe Azure data warehouse would be on top of my list for platform.

author avatar

That very much depends on your requirements. What’s the usage profile of your planned data warehouse solution? Your use cases drive the evaluation and the criteria used to weighted-rate aspects of potential solutions.

- Stefan

author avatar

Scalability, flexibility, performance, and ease of management are key. While a traditional RDBMS (originally designed for OLTP) can perform Data Warehouse tasks, you will always be better off with a purpose built platform that was designed from the ground up specifically for Data Warehouse data loading/ingest, report processing, and ideally in place analytics. Highly flexible, cloud-based solutions have reached a level of maturity and a price point that are extremely compelling - take a look at Snowflake.

author avatar
Real User

The type of data warehouse platform!

author avatar

The scalability of the data warehouse and its ability to handle multiple queries at the same time without loading the servers.

Find out what your peers are saying about Snowflake Computing, Oracle, Micro Focus and others in Data Warehouse. Updated: September 2021.
540,884 professionals have used our research since 2012.