MongoDB vs Snowflake: Which is a better fit?

Speaker 1: Hey, I'm Greg. I am an operations fellow at an entrepreneurship center. And I give MongoDB a four out of five. So we switched from Google Sheets, we were using it to store a lot of our membership data, but we found that it was just too unorganized and it wasn't able to handle the amount of data we were processing. We chose MongoDB for a variety of reasons, the primary one is that our staff rotates pretty frequently just as a part of the program and MongoDB is really easy to learn and understand. It's customer friendly for developers, but also it is able to be used by people who don't necessarily have database experience. We found MongoDB to be pretty easy to integrate with our apps. And while there are other services out there that provide similar functionality, the free tier on MongoDB was really a benefit to us and provided us with enough support to get started on the platform, but allows us to grow if we need it in the future. MongoDB was relatively easy to integrate with our system. We were lucky in that we were overhauling a lot of our backend data processing stuff anyways. So MongoDB was just one extra step to add on top of our existing code base. It was relatively easy to add different models and different ways of storing data. And we found that compared to things like SQL that may have a higher learning curve, because MongoDB's very user-friendly experience, we were able to do so without much time committed to learning, which as I mentioned earlier is really great given how often our staff rotates. So it really made the whole process a lot easier and it was one less thing that we had to think about as we were shifting away from our old model. So if you're thinking about trying out MongoDB for your business or service or project, I'd recommend messing around with their playground option. Their free tier is more than enough for a lot of smaller projects and it gives you a chance to learn MongoDB, all the ins and outs and see how easy it is to integrate into your system. I had some questions that were quickly answered through their online programs. But more than that, just being able to use it in my own application, just trying it out. That said, if you are considering MongoDB, look at Google Firebase as well. They have a database, I think it's called Firestore, as a part of the Firebase platform as a whole, but it integrates really nicely with all of the other Google services that are part of the Firebase platform. And what's on top of that, it still uses the same database structure and that all objects are stored as JSON objects. But it integrates with all of their other services pretty well.

Speaker 1: Hi, I'm David, Head of Engineering. I give Snowflake a five out of five, and for more reviews like this, click below. Before using Snowflake, we were using a combination of different clouds and different databases. So we had data in SQL Server in Redshift and MySQL across AWS, across Azure on premises. So we were looking for a solution that allowed us to aggregate all of our company data and product data in one place that was accessible to all of our people all the time. We chose Snowflake because we felt it had a great range of features. So in particular, the ability to push security right down to the database level, via role level security, via data masking, but also as well the ability to run different workloads at different combinations of CPU, RAM, and memory without having to size everything up to the maximum possible that we might need. So we thought Snowflake had a lot of versatility and a lot of great features built in. Getting Snowflake into our business was straightforward. The Snowflake team absolutely helped us. We had an account team who worked with us and gave us knowledge and training, but even say there were some learnings, we had some difficulty initially working out how we could migrate data, particularly not just an initial load, but ongoing changes into Snowflake from various sources. Also as well, Snowflake, one thing you really have to be across it is how it's priced and how it's costed because it does charge you on consumption based on the type of workloads you're running. So there was some effort on our part to work that out, to tune that, to try to make sure we are controlling costs. My advice for somebody thinking about getting Snowflake or getting started with Snowflake is to think carefully about the type of application you want to use Snowflake for, because your costs will absolutely be directly related to how you use Snowflake. Also my other recommendation is to think about your security carefully, because Snowflake offers a lot of features out of the box to allow you to push security down to the database level through role-based access, role level security, data masking, and a range of other things like secure sharing. So design your security model carefully because you really have the opportunity to enforce it directly at the database level. So they are my two recommendations if you are thinking of using Snowflake.