Apache Spark vs. Matillion vs. Snowflake

Well suited: To most of the local run of datasets and non-prod systems - scalability is not a problem at all. Including data from multiple types of data sources is an added advantage. MLlib is a decently nice built-in library that can be used for most of the ML tasks. Less appropriate: We had to work on a RecSys where the music dataset that we used was around 300+Gb in size. We faced memory-based issues. Few times we also got memory errors. Also the MLlib library does not have support for advanced analytics and deep-learning frameworks support. Understanding the internals of the working of Apache Spark for beginners is highly not possible.

Incentivized

Great: Need to query simpler APIs, or utilize well known services such as GSheets etc.? Matillion has got some of the best and easiest to use connectors out there. Not so great: Do you need have a competent CI/CD flow that you will be able to update / compare from Matillion as well as other sources at the same time? Good luck, you will need to be extra careful, as you might have to have a deeper dive into your servers Terminal each time you have a git conflict.

Incentivized

Snowflake is well suited when you have to store your data and you want easy scalability and increase or decrease the storage per your requirement. You can also control the computing cost, and if your computing cost is less than or equal to 10% of your storage cost, then you don't have to pay for computing, which makes it cost-effective as well.

Incentivized

• Rich APIs for data transformation making for very each to transform and prepare data in a distributed environment without worrying about memory issues
• Faster in execution times compare to Hadoop and PIG Latin
• Easy SQL interface to the same data set for people who are comfortable to explore data in a declarative manner
• Interoperability between SQL and Scala / Python style of munging data

Incentivized

• The user interface of your data pipelines makes it easier for people who aren’t as techy as data engineers to observe what's going on.
• Customer support is quick, not always as efficient as you would want it to be, but still.
• Nice documentation available.

Incentivized

• Snowflake scales appropriately allowing you to manage expense for peak and off peak times for pulling and data retrieval and data centric processing jobs
• Snowflake offers a marketplace solution that allows you to sell and subscribe to different data sources
• Snowflake manages concurrency better in our trials than other premium competitors
• Snowflake has little to no setup and ramp up time
• Snowflake offers online training for various employee types

Incentivized

• Memory management. Very weak on that.
• PySpark not as robust as scala with spark.
• spark master HA is needed. Not as HA as it should be.
• Locality should not be a necessity, but does help improvement. But would prefer no locality

Incentivized

• Matillion is brilliant at importing data -- it would be amazing to have more ways to export data, from emailed exports to API pushes.
• Any Python that takes more than a few lines of code requires an external server to run it. It would be great to have more integration (perhaps in a connected virtual environment) to easily integrate customized code.
• Troubleshooting server logs requires quite a bit of technical expertise. More human readable detailed error handling would be greatly appreciated.

Incentivized

• Add constraints for views and not just for tables
• Do not force customers to renew for same or higher amount to avoid loosing unused credits. Already paid credits should not expire (at least within a reasonable time frame), independent of renewal deal size.

Incentivized

Capacity of computing data in cluster and fast speed.

With the current experience of Matillion, we are likely to renew with the current feature option but will also look for improvement in various areas including scalability and dependability. 1. Connectors: It offers various connectors option but isn't full proof which we will be looking forward as we grow. 2. Scalability: As usage increase, we want Matillion system to be more stable.

Incentivized

SnowFlake is very cost effective and we also like the fact we can stop, start and spin up additional processing engines as we need to. We also like the fact that it's easy to connect our SQL IDEs to Snowflake and write our queries in the environment that we are used to

Incentivized

If the team looking to use Apache Spark is not used to debug and tweak settings for jobs to ensure maximum optimizations, it can be frustrating. However, the documentation and the support of the community on the internet can help resolve most issues. Moreover, it is highly configurable and it integrates with different tools (eg: it can be used by dbt core), which increase the scenarios where it can be used

Incentivized

We are able to bring on new resources and teach them how to use Matillion without having to invest a significant amount of time. We prefer looking for resources with any type of ETL skill-set and feel that they can learn Matillion without problem. In addition, the prebuilt objects cover more than 95% of our use cases and we do not have to build much from scratch.

Because the fact that you can query tons of data in a few seconds is incredible, it also gives you a lot of functions to format and transform data right in your query, which is ideal when building data models in BI tools like Power BI, it is available as a connector in the most used BI tools worldwide.

Incentivized

1. It integrates very well with scala or python. 2. It's very easy to understand SQL interoperability. 3. Apache is way faster than the other competitive technologies. 4. The support from the Apache community is very huge for Spark. 5. Execution times are faster as compared to others. 6. There are a large number of forums available for Apache Spark. 7. The code availability for Apache Spark is simpler and easy to gain access to. 8. Many organizations use Apache Spark, so many solutions are available for existing applications.

Overall, I've found Matillion to be responsive and considerate. I feel like they value us as a customer even when I know they have customers who spend more on the product than we do. That speaks to a motive higher than money. They want to make a good product and a good experience for their customers. If I have any complaint, it's that support sometimes feels community-oriented. It isn't always immediately clear to me that my support requests are going to a support engineer and not to the community at large. Usually, though, after a bit of conversation, it's clear that Matillion is watching and responding. And responses are generally quick in coming.

Incentivized

We have had terrific experiences with Snowflake support. They have drilled into queries and given us tremendous detail and helpful answers. In one case they even figured out how a particular product was interacting with Snowflake, via its queries, and gave us detail to go back to that product's vendor because the Snowflake support team identified a fault in its operation. We got it solved without lots of back-and-forth or finger-pointing because the Snowflake team gave such detailed information.

Incentivized

We were able to control on access and built various enviroment for implementation

Incentivized

Spark in comparison to similar technologies ends up being a one stop shop. You can achieve so much with this one framework instead of having to stitch and weave multiple technologies from the Hadoop stack, all while getting incredibility performance, minimal boilerplate, and getting the ability to write your application in the language of your choosing.

Incentivized

Fivetran offers a managed service and pre-configured schemas/models for data loading, which means much less administrative work for initial setup and ongoing maintenance. But it comes at a much higher price tag. So, knowing where your sweet spot is in the build vs. buy spectrum is essential to deciding which tool fits better. For the transformation part, dbt is purely (SQL-) code-based. So, it is mainly whether your developers prefer a GUI or code-based approach.

Incentivized

I have had the experience of using one more database management system at my previous workplace. What Snowflake provides is better user-friendly consoles, suggestions while writing a query, ease of access to connect to various BI platforms to analyze, [and a] more robust system to store a large amount of data. All these functionalities give the better edge to Snowflake.

Incentivized

We're using Matillion on EC2 instances, and we have about 20 projects for our clients in the same instance. Sometimes, we're struggling to manage schedules for all projects because thread management is not visible, and we can't see the process at the instance level.

Incentivized

• Business leaders are able to take data driven decisions
• Business users are able access to data in near real time now . Before using spark, they had to wait for at least 24 hours for data to be available
• Business is able come up with new product ideas

Incentivized

• Matillion has been the backbone of my company's analytical functionalities for 10+ years, so it has a good ROI.
• The price is ok for what our company built with it, but it starts to be less competitive if the tool is not used at its fullest.

Incentivized

• With separate compute and storage feature, the queries get executed quickly and it improves our overall productivity.
• Earlier we were using a different product for analytical purposes, but with Snowflake's in-built analytical feature we are now able to save money.
• Snowflake is cost efficient, features like auto suspend for compute resources helped to control the costs.

Incentivized