When people are interested in BI, professionals think of the word snowflake, then this thought should be in the foreground. But this article does not describe the aforementioned type of modeling – that would be almost too easy. We’re talking about the company and the product “Snowflake”, a very exciting cloud-based data warehouse solution. Snowflake is a real alternative for classic in-house data warehouses on your own hardware.
Snowflake’s Cloud Data Warehouse solution is a Software-as-a-Service (SaaS) solution. There is no hardware (virtual or physical) or software to select, install, configure or manage. The maintenance, the management of the hardware and software as well as the optimizations are taken over directly by Snowflake. The optimization of the queries, for example, is supported by a well-thought-out engine. As a Snowflake customer, you can save yourself from expensive data solutions where the data is not widely spread.
The principle of Snowflake is relatively simple: Load data into Snowflake and start the analysis straight away. These analyzes can be carried out directly in Snowflake or they can be approached with an external BI tool such as Power BI.
Strengths of Snowflake
According to Snowflake, their cloud-based data warehouse is characterized by the following strengths.
- You only pay for what you use effectively instead of an old-school data warehouse “on-premise” still costs the customer money, even if you are not actively using it. Examples of this are maintenance costs or process monitoring.
- Snowflake was designed for high-performance requirements. The patented architecture separates computing capacities from storage capacities. You can adjust the components at any time so that you get the currently desired performance. The data can be loaded into Snowflake at high speed and analyzes can be carried out at the same time.
- All your data is available together – the solution can handle structured and semi-structured data such as XML or JSON. These can be evaluated in the same warehouse.
- It is a complete SQL data warehouse – you are also able to query semi-structured data with SQL.
- Concurrency is a great strength – several users or groups can access data at the same time without reducing the performance of the queries.
- Data can be shared with third parties – warehouse data can be shared internally or even with third parties, e.g. a supplier. The third-party costs then also apply to the third party.
- Time travel as an innovative feature – Imagine that as a database administrator you have just accidentally deleted an important table or even an entire database. In today’s practice, you would probably have to work extensively with backups and restores. Snowflake enables the user to immediately go back to the «past» and retrieve an old database. This is how you can undo your mistake. Snowflake also knows the undrop command at this point.
Many conventional architectures combine memory and computing power. India Snowflake Companies, however, separates these from one another and thereby enables automatic, quick, and easy scaling of the memory, analyzes or resources for each job.