Data Migration: How to Get the Most Out of Snowflake POC
16.02.2021Decisions about your future data platform are not trivial. A data platform is like the foundations of a house. Once laid down and you start building on top it is hard to change things without much loss. Therefore, you must choose wisely. Before you make the decision, you should experience a potential data platform first-hand, on your sets of data.
POC Objectives and Scope
When defining your proof-of-concept have in mind what you would like to achieve. POC use cases should be aligned with the goals you would like to reach with your future solution.
It pays to pick scenarios that are either the biggest pain points of the current system or the ones you see as the biggest opportunities. POC is worth spending some more effort on because at the end, you should have a clear Yes or No answer on whether the tested platform is capable of supporting your business and technical requirements. If the POC is thoughtfully set, nothing will be lost. You will be able to reuse the already built procedures during the real migration.
During the evaluation process, you should understand the value that a new data platform can bring by:
- understanding key concepts and technologies
- testing the user experience
- learning about its killer features
- having concrete comparisons in capabilities between the current and new system
There is much more you will gain with a proof-of-concept. Tangible results will bring higher confidence and motivation among stakeholders and users. A side effect that cannot be ignored.
A Practical Example
Let’s say that you found yourself in a situation where:
- Your existing system is running out of resources, which are at the same time very expensive.
- Business users are starting to complain about getting their daily reports being way too late. It is past noon before data is refreshed and on top of that, it takes a long time for crucial reports to return data.
- As a result, the business is beginning to suffer.
You identified significant performance improvement as the must-have of the new data platform and you would also like to establish the grounds for an easily scalable platform embracing advanced analytical capabilities and new business opportunities.
As such it is highly recommended that you try at least to:
- Test one of the queries with performance issues on the new platform.
- Run one of the slowly working reports directly from the new database.
- Get familiar with advanced capabilities that the new solution has to offer.
- Sketch a brief roadmap of how you could expand the system after the migration is completed.
In order to test the performance, you will need to migrate relevant database structures, gather and import the applicable data sets, migrate the load procedures and pinpoint your BI tools to the new database. Depending on the legacy systems some minor code adjustments are expected.
Testing Snowflake Data Cloud
To get yourself familiar with the new platform, you should try some of its features first-hand. When you are accessing the Snowflake Data Cloud, you can choose from many options.
For starters, you absolutely need to check for yourself Snowflake’s ability to scale to any volume of data and any number of users. May we add a little warning here. Snowflake can turn out to be blazingly fast in comparison to your old setup. As a result, when organizations start testing some of their queries that require hours, maybe even days to return the results, they feel perplexed when suddenly that same thing happens in just seconds. Therefore, the first response we get is – it doesn’t work. But it does and this is where its greatness just starts.
Furthermore, you can utilize some semi-structured data. You can create multiple data copies by using the simple “clone” command. Access historical data with the “time travel” feature or see how easy it is to automatically ingest streaming data with Snowpipe and make it immediately accessible to the users.
And last but certainly not least, you should definitely not skip learning about the modern way to share and monetize your data with Snowflake’s data sharing capabilities. By using them, you can empower your insights with the data sources available on the Snowflake Data Marketplace.
Free Snowflake Data Cloud Test Drive
To make it easier for you and to dispel any doubts, Snowflake offers a free trial. To maximize the value of your proof-of-concept, we can:
- Help you identify meaningful use cases to test and find opportunities to grow the future solutions.
- Train you in the use of the new platform.
- Minimize your investment in POC and accelerate the implementation with our migration assets. You have to participate in the discovery workshops and the preparation of the relevant data sets, but from there on out we can take care of the rest.