Thanks for being this week with us!

I think learning about Snowflake makes great sense for somebody who has some on-prem DWH experience but not so much in the cloud (but wants to branch out). If you understand the basics of DWH and are proficient in SQL, you should be ready to start learning Snowflake.
If you want some knowledge of cloud data warehousing before you start learning Snowflake, there is a good resource that I’ve found that will only take you a few hours to digest: https://resources.snowflake.com/ebooks/cloud-data-warehousing-for-dummies
It is, of course, going to be a bit biased toward Snowflake as the choice for a cloud data platform but the content is really good and easy to read and understand.
Best of luck to you Philip Dießner (author) and Brent Brewington (interested party)!

Thanks a lot for the detailed answer! That sounds like a very good entry point, I will look into it.

Liliana, you’re welcome! It has been my pleasure.
I have had several people tell me that my book has been helpful in passing the Snowflake core certification exam. That said, I’d recommend you take a look at the official Snowflake study guide to make sure you are focusing on the necessary topics to pass the exam. Your first stop should be: https://www.snowflake.com/certifications/ where you can then find details about any of the Snowflake certifications you want to know more about.

Whenever someone instant messages me on LinkedIn with a very specific technical question such as this, I point them to the Snowflake forums. That way, the questioner can get the benefit of responses from the entire Snowflake community to help solve unique issues. And you can also interact with the Snowflake product experts who monitor the forums; it’s important that they know what questions users are having so that they can improve documenetation and/or add new functionality that is frequently requested but not yet available. I encourage you Shalltear to get involved with the Snowflake community to get answers to any specific technical questions such as this and to also help answer questions that others may have for which you have the answers. You can participate in the community forums at https://community.snowflake.com/s/

Thanks. Spoke to Snowflake directly this morning. Turns out, right now that’s not possible. They are working on it.

Great question Low Kim Hoe! In early days (2015), Snowflake’s original positioning was “The Data Warehouse Buiilt for the Cloud”. While Snowflake was hugely successful with its data warehouse offering, there was so much more that was possible with Snowflake beyond data warehousing. Thus, in 2019, Snowflake branded itself as THE “Data Cloud”. This rebranding helped call out Snowflake’s ability to federate the data and the fact that Snowflake was built from the ground up to be a data sharing platform (the actual workload is now called the “Data Collaboration” workload). Therefore, to answer your question, it is preferred to use the term “Snowflake Data Cloud”.

Noted on this. Thanks for your speedy reply!😁

That is a really tricky question to answer, Laura! Snowflake and Databricks are both great products, depending on the use case, so I don’t think there will ever be just one “winner”. And it is not unusual to see clients who have both Snowflake and Databricks, for a variety of reasons. However, I have a preference for Snowflake (probably no big surprise to you!) over Databricks. The general consensus, and what I’ve experienced, is that Snowflake is easier to get up and running right away and overall costs are lower with Snowflake when you consider the specialized (and costly) skillset needed to implement and manage a Databricks project.

As far as the future, I think both Snowflake and Databricks continue to evolve so I don’t know that we’ll see a return to their original areas of expertise. But it will interesting to revisit this conversation again in a year (or two!)

A little bit of an off topic question, but do you have any advice for data professionals/folks with more of a technology focused background to get a better understanding the business value of data and be able to speak more to that side? I don’t know if there’s an easy answer to this beyond getting an MBA/or becoming a CPA/doing career change, but was curious to get your take on this. 🙂 Thank you so much

Actually, I can provide you an on-topic answer to the question Stacey Kellough!
One thing I’d recommend is to read about some of the trends about the business value of data that you find interesting. For example, I’m really excited about Data Monetization. Snowflake has some really great resource materials on the topic, especially because of Snowflake’s advanced data collaboration (data sharing) features which make data monetization possible.
Here are a few Snowflake resources for Data Monetization that I’d recommend:
(1) Introduction to Monetizing Data: https://www.snowflake.com/guides/monetizing-data
(2) Modern Data Monetization Strategies:
https://www.snowflake.com/resource/modern-data-monetization-strategies/
(3) Blog post articles written by Jennifer Belissent

The one really nice thing about an electronic copy is that you can do quick and easy searching. However, I, too, really like having a copy to put my hands on and to refer to Julia!

As far as the number of hours spent working on the book, I didn’t keep track. Not sure exactly why I didn’t but maybe because it was something that I was doing that I was really passionate about and not something that I was counting on for the financial compensation. Maybe I just didn’t think it would take me as long as it did so I didn’t track my hours in the beginning. Not really sure.
Interestingly, in the beginning, my O’Reilly editor and I agreed on a schedule and I thought that there was no way it was going to take me 15+ months to bring an O’Reilly book to life and boy was I wrong. Initially, I was ahead of schedule but lots of things happened. For one thing, I originally wrote the book using the classic UI because the new Snowsight UI was generally available yet. So, then I had to go back and rewrite major sections to incorporate Snowsight. I just couldn’t, in good conscience, leave things as is with the old user interface. I’m really glad that I took the time to do that but it meant that I finished the book exactly on schedule, not before.
As a first-time author I had no idea everything that was involved. My editor and I laugh about it now. I’ve just started work on my second book and I have a much more realistic idea of what it’s going to take to get it across the finish line this time.

Awesome😊👍 I agree with you, if passion about what are you doing - you are not counting time for it.

Yes, Julia, the O’Reilly Snowflake Definitive Guide was my first book! But, it’s certainly not my last one! I’ve signed up to write the O’Reilly book titled “Hands-on Salesforce Genie” which (fingers-crossed) will be on bookshelves the early part of 2024.

Excellent question Buvan
There is no one right answer to this question. Without knowing more details about your architecture and role requirements, I’ll provide a high level response with some suggestions.
There are things you can use within Snowflake to (1) help streamline and optimize role management, and/or (2) processes you can put in place, and then (3) there are also tools that you might want to consider.
You might want to consider designating some databases as having open access to make it clear which data, if any, is available to all. While this prevents interruption to data consumers’ access, it could pose a security and compliance risk.
Some organizations have had success in simplifying management by defining only a limited number of roles per most databases. One example would be to have a Read Only and a Creator role per database.
Using role hierarchies is also a great way to create an abstraction layer where roles can gain access to other roles’ privileges through inheritance. If you are not currently using role hierarchies, this is definitely one thing I’d recommend researching to see if it is appropriate for your use case.
When user / role management is an activity that requires a dedicated resource or team, you could consider using ServiceNow or Salesforce Case functionality where users can request new user accounts and/or roles. This also allows you to document requests and have an approval process in place.
Also, you could take a look at Dataops.live which can be used to help manage a large number of roles. See this article for more details: https://www.dataops.live/blog/security-considerations-ina-dataops-world

There are two reasons why I don’t think you need an extensive background in data engineering to get the most out of the book jjp.
First, in comparison to other data cloud platforms, Snowflake is cosidered to be a much easier platform on which to design and deploy. Even if you are not a career data engineer, you should be able to understand most (if not all) of the concepts in the book, including the data engineering topics.
Second, the book was purposely designed so that each chapter stands on its own. That way, if there is some specific knowledge you want (that is not related to data engineering, for example) or need right away then you can tackle that chapter first without having to complete any previous chapter exercises. If, for example, you have a need to learn about Snowflake administration (i.e., user and role management) then you can gain a good understanding of that material. Later, you can dive into Snowflake data loading and unloading or other data engineering related topics.

:thank_you:

The physical (paper) book goes up to page 442, including the index.

The book wasn’t really meant to provide a comparison between Snowflake and other solutions. The intent of the book is to inform and educate the reader about Snowflake, assuming the reader is someone who has already decided on Snowflake or who wants to know more about Snowflake to be able to decide if Snowflake is a good fit for them.
That said, there are many reasons why individuals and companies choose Snowflake. Data storage factors (i.e., the amazing Snowflake micropartitions, for example and the fact that Snowflake bills customers based on compressed data size) are one consideration. Other reasons include ease of use (powerful web UI capabilities, can access existing cloud data, time travel capabilities, simple role based access and more), and the existence of Snowflake workload capabilities such as data collaboration (data sharing) for which Snowflake is incredibly well-known for.
And it isn’t necessarily that Snowflake would be the only data storage solution selected although Snowflake supports data warehousing and data lake workloads, among others. In larger organizations, it is not uncommon to have Snowflake in addition to another different data storage solution. It really depends on the use case(s).

Lirone Perhaps a few other reasons why Snowflake could be a good choice, too, for those interested in AI / ML — Snowpark and Snowflake’s acquisition of Myst are other excellent reasons! 🙂

Excellent question Alejandro Marmol! Many of the clients I see getting started with Snowflake are looking to improve their insights so they are not necessarily at the beginning stages of their data journey. That said, I don’t think there is necessarily a right or wrong time to apply Snowflake.
The really great thing about Snowflake is that if the use case lends itself to using Snowflake, you can get started right away. For example, if you already have data sources that exist in the cloud (in an S3 bucket, for example), you don’t even necessarily need to move the data into Snowflake to be able to surface and use the data for insights right away in Snowflake!
Another specific example I frequently see is clients who have their logic built into a BI tool which locks them into that specific BI tool; they’re looking to extract that logic from the BI tool and put it into Snowflake instead which allows them to have more of a “Bring Your Own BI Tool” (BYOT) approach.
Because Snowflake implementation and maintenance rely primarily on having some SQL knowledge, no unique or hard-to-find skillsets are needed which means it is really a good choice to get started with at any data maturity stage.

Ohhhhh! That’s interesting!!!😄 I will really be glad to enjoy that book!!

You’re welcome. It’s been my pleasure to join you.
To answer your question, I’d recommend that you at least be familar with relational databases and SQL to get the most out of the book, Roman Zabolotin. In addition, to take advantage of the vast product offerings that Snowflake has to offer, you’d probably want to be knowledgeable about data warehousing and know a little something about data engineering or at least be working with a team who has those skillsets. Because some of the Snowflake workloads, such as Data Sharing / Data Collaboration, are likely going to be new to most everyone, most people probably won’t have the background in that workload so it will be a new topic for most everyone.

You may be referring to the Snowflake Marketplace (a subset of the Data Collaboration workload) where data providers can monetize their data by offering it to the public. There also exists the functionality to create a Snowflake private data exchange to monetize data with a set of known partners. Snowflake is definitely the leader in the Data Collaboration (Data Sharing) space.

https://www.snowflake.com/en/data-cloud/marketplace/ and https://www.snowflake.com/provider-policies/ are some good place to start with. I think you can provide data as a data producer, and it can be used by other customers by snowflake’s marketplace (it can be free or paid)… Example: https://www.snowflake.com/datasets/starschema-covid-19-epidemiological-data/ they provided the covid data sets. So, the customers no need to reinvent the wheel, and they can concentrate on the business and these providers are responsible for the data. FYI

Yes Snowflake does offer a free 30 day trial – https://signup.snowflake.com/
Here is a tip for you Roman Zabolotin – Snowflake gives you a full 30 day trial where the time starts on the first day of the next month. So, for example, if you sign up for a free trial the first week of the month, you’ll get the remaining 3 weeks of the month plus the full next month. The free trial is that 30+ time period or $400 usage credits, whichever one comes first. So, you’ll want to be smart about how you use the Snowflake trial org so you don’t run out of usage credits. The amount ($400) is a large amount of usage credits so it’s unlikely you’d run out of usage credits but I just wanted to call that out. The usage credits start from the first day that you create the trial account and go through the life of the trial account.

Thanks, I think it’s a very useful trick.