What is our primary use case?
We mostly provide cloud data warehousing platforms for major banks in Canada. What we're trying to do is to create a standard platform environment that is compliant with the regulatory requirements imposed by the government and financial overseeing institutions for the banks. We help them to onboard the lines of business to these platforms and migrate the existing workloads to the cloud platforms.
What is most valuable?
The ability to scale out services on-demand and scale them down when they are not required is most valuable. You are in control of your expenditures, and you are also in control of the horsepower that you need. That's a major advantage.
What needs improvement?
The major challenge that we're seeing with Azure Synapse is around security concerns. The way it is working right now, it has Managed VNet by Microsoft option, similar to the implementation of Azure Databricks, which may pose a concern for financial institutions. For managed environments, the banks have very strict policies around data being onboarded to those environments. For some confidential applications, the banks have the policy to encrypt it with their own key, so it is sort of like Bring Your Own Key, but it is not possible to manage the resources with Microsoft or Databricks, which is probably the major challenge with Azure Synapse.
There should be more compatibility with SQL Server. It should be easier to migrate solutions between different environments because right now, it is not really competitive. It is not like you can go and install SQL Database in some other environment. You will have to go through some migration projects, which probably is one of the major showstoppers for any bank. When they consider Synapse, they not only consider the investment in the actual service; they also consider the cost of the migration process.
When you scale out or scale down your system, it becomes unavailable for a few minutes. Because it is a data warehouse environment, it is not such a huge deal, but it would be great if they can improve it so that the platform is available during the change of configuration.
For how long have I used the solution?
I have been using this solution for two years.
What do I think about the stability of the solution?
Generally, it is stable. We all heard about the Active Directory issue earlier this week, but it was not related to Synapse. It was related to the Azure platform.
What do I think about the scalability of the solution?
Its scalability is good. The only thing is that when you scale out or scale down your system, it becomes unavailable for a few minutes. Because it is a data warehouse environment, it is not such a huge deal, but it would be great if they can improve it so that the platform is available during the change of configuration.
How are customer service and technical support?
We're working with banks, and they have great support because Microsoft has multiple representatives closely monitoring each account. Whenever there is an issue, they're being proactive. They're making a lot of money out of it. Most of the banks, on average, spend between 30 to 50 million a year on Azure. They're pretty large accounts, and Microsoft has dedicated people supporting everything related to Azure.
Which solution did I use previously and why did I switch?
We deliver platforms to different banks. Some of the banks go with Synapse, and some of the banks go with Snowflake. Overall, these are two major alternatives available right now.
There are multiple differences in terms of the support of different workloads. When one of the banks made a decision to go with Snowflake, the major reason for it was the support for the multi-cloud environment. The major pro of Synapse is the service, and the major con is that when you decide to move out of Synapse, you would have to rewrite the entire thing, whereas, with Snowflake, it would be just simple migration to different cloud providers.
How was the initial setup?
It is relatively straightforward as long as you understand what you're doing.
What about the implementation team?
You don't really need to maintain it. That's the entire point of the cloud. You pay for it to be maintained.
We do deal with monitoring and other similar things, but most of the activities are automated. Overall, it doesn't require a lot of labor around it. We're delivering the platform as infrastructure as a cloud, so everything is going through the pipeline.
What's my experience with pricing, setup cost, and licensing?
It goes by the usage, and there are some limits. Synapse goes by particular pricing, and it is expensive. Both Azure Synapse Analytics and Snowflake are pretty expensive. They don't have standard pricing. They deal with each customer differently.
What other advice do I have?
For working with Synapse, you need to have an understanding and knowledge of the product to take full advantage of it. Synapse has a lot of features in terms of scalability, such as resource management, distribution, and partitioning. There are a lot of things that you need to consider when you go for it. It is not a simple database that you put in there, and it is running itself.
I would rate Microsoft Azure Synapse Analytics a seven out of ten.