AWS Lambda is a very popular tool for deployments at JBS. As an AWS Partner, we have repeatedly espoused the benefits of serverless architectures as...
What is Snowflake and why should I care?
Getting Into It
Many organizations are considering cloud-based data solutions, instead of planning out on-premise solutions. Organizations are also evaluating moving their existing on-premise solutions to the cloud. When making these considerations, there are a few options that are all quite good. The major players in this space are:
While each option has its own benefits and pitfalls, Snowflake offers some valuable benefits that make it a compelling option.
What is Snowflake?
Snowflake is an infinitely scalable, easy to use data platform. Snowflake also allows for storage and compute needs to scale independently of each other. Moreover, the data-sharing capabilities within Snowflake are some of the best in the industry.
Snowflake can be set up to leverage either Amazon Web Services (AWS) or Microsoft Azure which means that your existing choice in cloud platform shouldn't be impacted if you are on one of those platforms. Snowflake focuses on the data warehouse components of the solution and does so quite well. That being said, it is not all-encompassing. Things like the presentation layer as well as the Extract, Transform, Load (ETL) layer will be separate but can tie in nicely.
Benefits over AWS RedShift?
AWS RedShift is one of the tried and true players in this space and it’s typically a go-to that people think of first when they look to set up a cloud-based data solution. It’s hard to come up with downsides to RedShift but there are a couple.
One area where Snowflake comes out ahead is the cost structure. With RedShift, costs for storage and compute needs are coupled. With Snowflake, the costs for storage are decoupled from computing needs. Decoupled cost structures allow for your business to grow at rates that align with your needs, avoiding paying for things you may not need.
Another area that Snowflake shines when compared to RedShift is with initial setup time and effort. With Snowflake, setting up your data solution is much simpler. Snowflake abstracts the components such that the end-users can get up and go quickly. All the setup and maintenance is handled by the Snowflake architecture, taking that work off of the hands of users.
Lastly, RedShift is an Amazon product. As such, it is only compatible with AWS cloud infrastructures. An argument could be made for cloud vendor lock-in. It is abnormal for companies to switch cloud platforms. In any case, it’s good to have a full picture of all the options available as well as their benefits and pitfalls.
Benefits over Azure Synapse Analytics?
Recently the naming for Azure’s Data Warehouse switched over to Synapse Analytics. When comparing Snowflake to the old Data Warehouse, it was a much easier win for Snowflake. When Azure upgraded their data platform offering, they opted to allow for users to segment the compute and storage costs, just like Snowflake, so now they have parity in that area. Azure also now allows for pausing of computing for Synapse Analytics which makes it harder to find benefits between the two options.
That said, Azure’s offering is quite pricey. The typical cost ends up being a little more per month to the end-user. That cost is due to Azures' full suite of products complementing their data platform. If you are in the Microsoft vertical, Synapse Analytics is a much easier integration path of the “big three” offerings. In order to mitigate the higher cost basis, users can leverage reserved instances.
Storage costs with Synapse Analytics are segmented from the compute costs which is great. However, the tiers for the storage costs are quite large. When you are storing data, it goes in terabyte increments. So if you have a dynamic data set, this can be problematic since the storage costs will be higher due to the way the tiers are defined.
Another area where Snowflake wins over Azure Synapse Analytics is avoiding vendor lock-in. Snowflake gives your organization the freedom to select their cloud platform of choice and it will still work as expected.
Final Thoughts
When evaluating the options available for a cloud-based data platform, there is no bad option from the options listed. However, Snowflake does offer some valuable features which make it a compelling option for businesses looking to move their data platform to the cloud. Snowflake deserves to be an option when making the decision at your organization. However, if you are already engaged with a product such as AWS RedShift or Azure Synapse Analytics, you would have to weigh the pros and cons per your particular situation.
Snowflake does allow for you to try out their products with a free trial, so it’s worth checking out before making a selection for your organization.
