Snowflake Analytics Reviews

We use this solution to move on-prem data to a cloud data warehouse where we have a requirement for more computation storage or when we need a plugin to work with specific technologies like ML, AI and Data Science. 

The advanced features like time travel, zero copy cloning and scalability have been most useful. Snowflake requires zero maintenance for Data Warehousing on the cloud system. It is more secure compared to any other data platforms that we have used. The tri secret key is the feature that gives a lot of confidence to customers. 

Moving data from legacy systems to Snowflake is not that easy. There are some cases where processors are not actually compatible with Snowflake. If Snowflake could offer compatible versions of other legacy system objects, that would be more convenient for anyone to move to Snowflake. 

When comparing MySQL with Snowflake, MySQL is faster in executing the complex queries compared to Snowflake. This is the only factor where Snowflake lags behind competitors in the market.

I have been using this solution for four and a half years. 

This is a scalable solution.

The customer service and support for this solution is good. We had previously failed to retrieve data after the time travel period. We had to raise a ticket with technical support team and they helped us. They charged us for this but were able to help us retrieve it.

Positive

The initial setup is straightforward. The beauty of Snowflake is that you can select the configuration you need. For example, if you have a requirement of a one terabyte load with high availability and 10 users, Snowflake will set it up for you to match these requirements. 

The ROI you receive using this solution depends on your use case and if you are starting from scratch or migrating from a legacy system. We have experienced a 40 to 50% improvement in ROI compared to legacy systems.

When using Snowflake, you pay based on your usage. They calculate how much CPU has been used. If you use excess warehouse storage, you are charged one credit per hour. If you are in Asia, you are charged $3 per credit. If you have 10 users running parallel with the same excess, you will be charged $30. 

Pricing for storage is calculated in the same way. You will pay upfront for the storage that you know you will need to consume. There is this option of prepaid consumption as well as post paid consumption. If you pay upfront, you will charged $20 per hour, per terabyte. If you paying after use, you will be charged $40 per terabyte.

There is close to zero maintenance required for this solution. 

You don't need to use traditional warehouses, vacuuming or indexing. A database administrator or system admin is not required.

I would rate this solution a nine out of ten. 

Snowflake replaces enterprise data warehouses that have traditionally been on-premise and are now managed via the cloud. We are a services company and we create these products for our enterprise customers in a variety of industries including health care, education, travel and tourism. We supply this product and part of my job is to assess its suitability for my enterprise customers. I'm vice president of digital transformation and we are customers of Snowflake. 

The solution is completely managed which is great. You can create analytical products as a SaaS offering so nothing needs to be installed and it doesn't require any licenses. The solution can run on AWS, Azure and on Google Cloud and you can choose the underlying cloud on which Snowflake creates your databases, uses different services, stores your accounts and stores your data. It offers very good flexibility. It also offers the computation completely decoupled from the storage so you pay fees for computation cost only. If I'm computing for two hours a day but have to store my data permanently, I pay much less for storage. Performance-wise it's a good solution. 

If you have a lot of computations, it becomes very costly. There can be some complexity when designing your warehouses or enterprise data, but that is based on your requirements and not the underlying warehouse that you choose.

I've been using this solution for about three years.

It's extremely stable because it's backed up, load-balanced and has the elasticity that AWS provides. 

The solution is scalable and has the concept of creating virtual resources, so I can buy 10 gigabytes of data capacity, then decide to process that data every day, and it then goes into my storage creating multiple databases, and scales by itself. I don't need any calculations of load or throughput worked out beforehand. For most other similar products, such as RedShift or Azure Synapse, you have to work out your average load versus high peak load, and provision the warehouse. In Snowflake, you don't need to do that, it auto-scales up and down.

Customer service has an admin role, creating virtual warehouses and databases and establishing access. Whenever there's a change or a need for more resources, the Snowflake administrator has to provision for that. The other support is for when all the ETL pipelines and the data warehouse is established, changing anything requires ETL engineers. There's no support needed to manage the infrastructure.

The initial setup is straightforward. You buy an account, create a domain and the number of accounts you need, and that's it. It takes about five minutes. 

The solution can become expensive if you need to do a lot of computations. 

I highly recommend Snowflake wherever possible. 

I rate this solution eight out of 10. 

I mainly use Snowflake Analytics for Power BI reports.

Snowflake Analytics' most valuable feature is its inbuilt infrastructure for executing queries, which I don't have to manage based on my data volume as it's taken care of by Snowflake.

End-to-end execution of jobs isn't possible with Snowflake, which means we have to do some customization. Allowing sequential flow of orchestration would be an improvement. In the next release, Snowflake should include direct integration and communication between other cloud systems.

I've been using Snowflake Analytics for over a week.

Snowflake Analytics is stable.

Snowflake Analytics is scalable.

The initial setup was straightforward, and I would rate the process as four out of five.

Snowflake Analytics is a little more costly than Azure.

Snowflake Analytics is more managed than other cloud offerings, which means I don't have to deal with anything happening in the backend. For those thinking of implementing Snowflake, I advise analyzing the cost, your volume of data, and the number of users. If you have a huge demand on these things, I would recommend choosing Snowflake. I rate Snowflake as seven out of ten.