How to reach the stars? Board on to a Cloud!8.06.2020
Dive into Snowflake Cloud Data Platform series, Part 3
In the last years many of the businesses around the world have prioritized data management and analysis as part of their main strategies. That comes with no surprise because data is growing at an exuberant rate. Optimizing this quickly growing collection, its storage and analysis is therefore one of the most important projects for most companies around the world.
Goals are many. Knowing their customers better, identifying emerging trends before they happen, uncovering untapped market opportunities, improving internal processes for greater productivity at lower costs. When you examine the many benefits, you can really understand how data indeed is “oil of the 21st century”.
The results of this digital transformation surely are powerful. But on the way to the stars many questions arise. How to reach those shiny lights that will take you to the next level?
To understand the digital, it always helps to make comparisons with the real world. Imagine you’re a retail company, with a long tradition of selling bikes. Your great grandfather started selling them at the beginning of the 20th century. For a long time, he was selling one kind…and one kind only. Then, the lady’s version came, followed by a smaller framed bike for kids. Still, business was easily managed. One supplier. One storage. Simple layout – the bigger ones he kept on the left and the smaller ones on the right.
But then, evolution happened. Racing, city, mountain, e-bikes. Gear expanded. Bike parts, pumps, tools, bags, safety equipment, and the list goes on. One brand became 150 brands. Now imagine you would be still running that from that one simple storage from the beginning of the 20th century. Big ones on the left, small ones on the right. See where we’re going?
So, let’s translate this into digital.
- Bikes and gears are your data. Where are the times of simple excel sheets and some plain SQL. Now you get every day a batch of JSON, XML,… before you even open your eyes. Where do you keep them?
- Suppliers are your sources. For a long time, they were only one kind, produced in-house. Now, you’ve got the ERP, CRM, IoT, Third party, etc.
- That old storage is your database. It might as well be your old data warehouse.
As you might have guessed your first step on your digital transformation is organizing this storage. So, first things first – you need a modern data warehouse. And to make this project successful, you need to choose the appropriate one.
Cloud vs. On-premise Data Warehouse
Companies are debating – should we choose the ‘established’ on-prem solution or are we considering cloud? Is anything left that is keeping us on our legacy warehouse or are we ready to embrace the cloud warehousing? And, very common as well – our legacy is full. What next?
Let’s start with similarities. Both look the same. Both are big (some more than the other) chunks of hardware used as a repository of current and historical data from one or more disparate sources. And the main difference between the two is quite obvious one – legacy you own and the other you rent. Therefore on-prem data warehouse servers are located onsite at your organization and cloud servers are scattered around different locations and it makes most sense to opt for the location nearest to you.
Here are six simple questions to ask:
- How much time do we have? – ADMINISTRATION
Deploying conventional warehouse is a project that can take months. Cloud administration takes minutes.
- How much money are we willing to spend? – COST
With legacy warehouse, expect high initial investments – you’ll need to buy all the hardware you need up front, licenses for software and solutions, staff to configure and manage the system.
Cloud reduces large capital expenditures as it provides a pay-as-you-go model.
- What are our sources and what kind of data do we operate with? – STRUCTURED vs. SEMI-STRUCTURED
Today you can’t escape the semi-structured data. Cloud environment is more up to date and has more support for different data types (json, xml, parquet,…).
Let’s return to bikes for brief 8 minutes. For some hands-on magic, check out this video, where you can see how weather impacts bike rentals in NYC. You will learn how Snowflake Cloud Data Platform supports different data types and how you can manage them with simple SQL queries.
- What’s our data growth rate? – SCALABILITY
It is said that 90% of the existing data was created in last two years. So, no matter how big your legacy is, we can easily assume it will be full faster than we would expect it today. Scaling up on-prem systems is a time-consuming and resource-intensive task, as it usually entails purchasing and installing new hardware.
Bear in mind that hardware and compute resource often need to be over-dimensioned to cover spikes, which happens only 20% of the time. In long term hardware resources are not utilized enough which is expensive for every organization to maintain.
Snowflake Cloud Data Platform eliminates that need entirely, making scaling up or down accomplished easily and effortlessly. Unlike the traditional setups where constant upgrades and maintenance is required, cloud computing provides unlimited storage to ensure businesses never have to worry about increasing amounts of data.
- Is AI part of our business? – IMPLEMENTING AI
Today there are a lot of cloud solutions, that make implementing AI much easier – you don’t need to pay so much of data scientists to start creating a lot of models – a lot of set-up are already created for you. This availability made a lot of companies making the move.
- Last but not least – SECURITY
It’s still a lot of concern over the security of data stored in the cloud rather than on-prem, which is why you might be surprised to hear that more and more companies are beginning to see cloud computing as a more secure environment than hosting on local machines. Hardware gives the illusion of being more secure just because it’s a tangible reality- you see it, it’s there. But cloud data warehouse provider’s entire business model relies on data security and encryption. These companies invest heavily in security technology and dedicate entire departments to the protection of your data.
Snowflake stands out, as it as well has the data encrypted automatically. It’s a multi-tiered encryption of keys encrypting keys, encrypting keys… a very sophisticated process one can hardly believe until he sees it. To learn more about it read our interview with Kent Graziano, Chief technical evangelist for Snowflake.
We are certain most companies will benefit greatly by deploying Snowflake Cloud Data Platform because of the above-mentioned qualities.
If you’re eager to learn more, read about it in our next post How Snowflake tackled limitations of legacy data warehouses.