SQL guys : why don't we use CDS ?

[FredericForest]

Hey folks,

Having been more involved on these forums for the past few weeks, I realized there are quite a few SQL Guys like me around. (I thought I was an oddity)

Until recently, there was a gentle "live and let live" attitude between SQL and CDS. We didn't get much love (see the ideas on triggers or procs) but we were more or less free to do our own things in a nice cost-efficient manner. Now, with the new licensing, the attitude has switched to a much less subtle "convert or die" message... with the expected pushback from the community as a result.

Yet some folks at MS are curious about why we won't use CDS as our go-to data storage. I have my own reasons, which I'll share in the next post, but I'm curious about yours. Are you able to explain why you simply won't switch to CDS overnight? Obviously, I don't believe the "all-in on CDS" strategy is going to change anytime soon. The goal here is to provide insights and understanding to MS as to why we "resist" CDS (a genuine question I've been asked by different MS employees already).

To temper my comments, let me start by saying that I DO believe that CDS has a part to play in the Power Platform. It's a great low-code DBMS sitting somewhere between Sharepoint lists and full Az SQL DB as a backend option, not unlike front-end sophistication can vary between highly managed model-driven screens all the way to PCF components. As far as I can tell, CDS' main value proposition is : 

• No need to learn SQL (not an advantage to us, but will be for some)
• Common model with Dynamics (if your client uses that)
• Native REST API querying, something vanilla Azure SQL DB doesn't offer

@PaulD1 @MrMarshall  @MagnusGöransson @WillPage @OneThing @NEPatton @Rick72 @AngryBatVoice @sscarcella @digiservice @opit @JohnP @inverse70 @wjhepworth 

[FredericForest]

Now, my own two cents on the topic.

First, one has to realise that onboarding a new technology is a cost-benefits analysis. If we don't do it, it's because the benfits aren't justified by the costs.

Benfits

I briefly touched on the fact that, overall, CDS currently lacks clear benefits for me. I'm well-versed in SQL and my clients don't use Dynamics. Proper modeling is so core to a good project that I won't just go for something generic unless I have a compelling reason to. And creating 5-10 tables (an average for a PowerApps project) is just not that long. So the argument of the "one-size-fits-all" of modeling just doesn't hold for me.

Costs

The list here is a bit long, so let me break it down.

Learning

Any new technology needs time to learn and to master. We need to develop best practices and good patterns for said technology. There was nothing broken with SQL for us to start with, so even if we would have feature parity between the two platforms, the learning costs alone would have me stay in SQL. It's not that major in the big picture, but it's worth keeping in mind.

Licensing fees

Lots of ink has been spilled over this, but there's a HUGE price difference in the two technologies. Most of my PowerApp project work just fine on an Azure SQL DB S1 (30$/month) with no extra storage since we already have 250 GB for that price. By comparaison, a CDS-based project is 10$/month/user + a wooping 40$/GB/month (vs 0.20$/GB/month for extra SQL storage).

Consider an average project of 50 users with a 3 years lifecycle (lets say no extra storage is needed), the price difference is 17K$ ! This alone kills my sales argument to client that low-code is great because it's faster/cheaper. Right now, full-blown custom dev by a senior consultant is cheaper... And that's a problem ! ^^

But enough has been said on this already.

Physical layer

Part of our job is making fast performing apps, and a big part of this is optimizing the physical layer of data. There's a big difference between a TINYINT and a BIGINT, between a VARCHAR and an NVARCHAR. And if you join tables on GUIDs, we're just not going to be friends. I need to place an index here and there, sometime add additional fields to cover a query and occasional materialize an entire view when it makes sense for the user experience. To the best of my knowledge, we don't have any control over that in CDS.

Complex querying

CDS views are nice, but I often need much more complex logic to represent common business problems. I need the expressivness of the T-SQL language for things like windowing clauses, pivoting data results, CTEs and paging with OFFSET/FETCH just to name a few. And I often need to upserts batches of lines at once with MERGE statements from a parsed JSON payload.

SQL programatic objects

I need strong T-SQL procedure support with transactions enabled, elegant error handling (beyond just bubbling up a system error) and dynamic SQL. On rare occasions, I need to temporairl elevate priveliges during a procedure in a controlled manner. Sometimes I need to materialize intermediary results in temporairy tables and the few times I've had to use CROSS APPLY on a table-valued function really saved my bum.

I understand we can add C# plug-ins to CDS, but my language is SQL...

Data Ops

Bringing CI/CD to a database project is quite the challenge ! Not only do you have to have deal with schema comparaisons, but you also have to deal with transitional changes of the data itself (it's a stateful system after all!) with pre/post deployment scripts.

CDS is improving in this area with solutions and CLI commands, but we're still worlds away from git-based SSDT projects with peer-reviewed pull-requests over an Az DevOps pipeline with automated testing of SQL programatic objects.

Connectivity

My projects don't live in isolation. We do BI on them, we exchange information with other systems and perform other various ETL work. Azure SQL DB inherits the hundreds of native SQL Server connectors to the outside world whereas CDS is still quite poor on this area. Power Query is nice enough to import data, permform light ETL work (more like data prep) and its got a great UX, but it doesn't cut it for serious industrial data pipelines. And yes, my PowerApps project sometimes import gigabits of data into their databases (that tend to be the case when we Papp as a light MDM/DQM tool for data stewards).

That's about the major parts for me. And you ? How do you feel ?

[wjhepworth]

For me, the deal killers for CDS came down to the following:

• The unecessarily complex implementation of relationships were very SharePointy and not very efficient.  I have not gone back and revisited.
• While it is great we start with a whole lot of entities and relationships, I did not need any of them.  The areas where it supported what I needed was generic and required me to extend them.
• Most Importantly - The Delegation limits were horrible!  Seriously, when you look at the delegation charts it seems like the CDS and SharePoint connector teams had to have gone on  vacation.  Simple function such as Average, Max, Min, Sum and IsBlank are all non-delgable on CDS (and SharePoint).  Compare that to SQL Server where nearly every function is delegable.  Simply put, the team that built and updated the SQL Server connecter are way ahead of all others.  No really, even the SalesForce connector is better.  

Just my $.02 after taxes.

-Bill
 

---

[AngryBatVoice]

The major one for me was just cost.  Feature- and function-wise, CDS might be the middle child between Sharepoint Lists and AzureSQL, but if that's its intended niche, someone forgot to inform the pricing people, because the license necessary for CDS was always just too far out of my reach to even consider it.  I'm a small team, working for a small company.  Just like with the new Powerapps and Flow licensing, even if I could use CDS to streamline and improve every single workflow in the company, we just aren't big enough to where those savings would justify the costs.  We simply don't have that kind of overhead. 

On the flipside, with SQL, I'd get a more robust product that I can leverage at an acceptable cost.

I honestly have not formed an opinion on CDS outside of the cost, because the cost eliminates it as a possibility for me.  I'd happily try it out, but at the price-point they have it at, it'd be a waste of time because it isn't something I'd be able to sell to the ownership.

[Anonymous]

@AngryBatVoice  @FredericForest  @wjhepworth  Moving to CDS board 

@Anonymous 

[OneThing]

More or less the same sentiments from us.

CDS would be great is all our systems were in Dynamics. The Built in entities would be easily shared and accessed across products and the ability to add the them would be useful.

However Dynamics change of prices had again prices itself out of our budget so there has been no further investigation. 

Almost all of our systems are independent and living on their own SQL databases. The ability for us to join these in AzureSQL and built upon the shortcomings of each platform is where our real benefit lies. 

[NEPatton]

My feelings on CDS?

I think it's awesome.  I like the concept and power it offers - for where it is the righ solution.  I could see how CDS is usable/valuable even game changing for some organizations/departments/businesses.

Why don't I use CDS?

I work in a manufacturing facility.  To build a modern manufacturing environment - you have to be able to connect equipment to the network and view the status of the equipment in real time and throughout history.  The equipment being connected is a huge hodge-podge - we have new equipment that we purchased within a few years, decades old equipment that we've retrofitted within the last decade, and stuff that we haven't been able to retrofit yet.  You have other non-equipment systems that you must interface with to effectively use your data:  Blue Mountain (Asset management software w/ SQL back-end), SAP, OSI PI DB/AF, MES, etc.

Long story short - I am trying to convert a manufacturing plant with over a hundred different data sources into (what appears to be, from the end user perspective) a single data source that allows for equivalent data comparisons/analyses across all hardware/software platforms, etc.  It is incredibly complicated to solve this problem - it requires capital expenses to strategically update hardware, it requires reprograming and redesigning equipment software architectures.  I have been working on this for 8 years now, and I still have a long way to go.

Anyone who tries to tell me that CDS can be a solution for what I'm doing (SUCCESSFULLY) at my plant does not understand what I do.  I have an open mind to new technologies - I am good at assessing the value of new technologies.  CDS is valuable!  So is snake venom! 

Valuable != Usable!

[digiservice]

Great question, it's just simply not feasible at the moment @FredericForest 

The amount of data we have in various SQL deployments it would be fairly close to impossible to get everything in to a new form without a mammoth project. We have data coming in from legacy systems, third party systems and internal developments. 

We don't have the budget, resources or time to move the data stores. And the cost and horror of getting suppliers to move away from SQL also. 

I think in a utopian scenario we'd all have our infrastructure completely different and we could see where CDS might fit in that picture. But the size and complexity of this kind of change for an organisation like ours would just not have the ROI. 

The big value for us and I'd guess a fair few others was that the SQL connector meant we could plug in to everything else we already had so fluently.

[PaulD1]

To this point, our clients haven't shown any interest in CDS (largely due to cost I expect) so we haven't looked into CDS deeply, though we have done some projects with canvas apps connecting to Dynamics which I believe is a similar experience. Below are my thoughts, but please be aware I know very little about Dynamics/CDS so happy to be corrected/proven-wrong.

Dynamics:

What I see in Dynamics (extended with additional entities) is some pretty horrible data modelling (same data in multiple entities) which I presume is because it is difficult to build performant queries on a normalised data model in CDS / Dynamics so data just gets repeated through multiple entities, though I stress this is speculation on my part.

Entities following the common data model have a gazillion fields - it is a real pain in PowerApps to figure out the correct one to use, especially as sometimes the internal name needs to be used and sometimes the 'freindly name'. Because there are so many unused fields, people start using them to store other data rather than create new fields which undermines the whole common data model approach.

In one project there was a desire to write data to a many-to-many relationship. This proved to be impossible (the child entity is invisible to PowerApps and Flow). We gave up and stored the data elsewhere.

When moving Dynamics from one environment to another (say Dev to Test) the environments frequently do not behave the same way - for example, one environment will want an entity use friendly names, the other internal names, one environment will extend out 'reference' fields (e.g. a UserID field will let you drill into the name, email, etc)., the other will not (e.g. returing only the ID). My knowledge of Dynamics is minimal, but I have seen this across several different customers whose Dynamics teams seem more than competent but are unable to get environments to behave consistently.

SQL

SQL is a know quantity, has great tools, is super powerful and flexible even supporting JSON and XML. There are good tools for tuning Views/Indexes to help you run efficiently and keep on the lower tiers of the subscription model. It is easy to separate out solutions into different DBs so it is easy to track of cost by App, do a restore from back-up by App if needed and so on. SQL DBs are easy to script out, so deploying to other environments is quick and simple.

Conclusion

In short, I'm much happier working in SQL and our customers haven't expressed interest in CDS. I guess CDS is easier/better than SQL for building business processes around the data, but so far for our customers that doesn't seem to be a compelling enough benefit to outwiegh the costs, though it is early days for CDS so that could change quickly.

[FredericForest]

Hey @PaulD1 ,
I get what you mean, I feel more or less the same way. I think the ease of navigating the CDM improves the more we use it, but the learning curve is certainly steep... and not very appealing if we don't need it outside of our PowerApps projects (which tend to be quite focused and specialized). Just a lot of noise for what we need, really. I does make sense however in a D365-based client were learning the CDM is part of the job anyway since there's going to be relevant data in it.

Moving to environments is also a big thing. I can't stress this enough : migrating from one environment to the next is not just about the schema, it's also about the DML scripts (pre/post deployment) to adapt the data.

And yeah, it's a real shame that you can't adress CDS using the SQL language (the de facto lingua franca of data), and T-SQL in particular (a nice version of it, which virtually all MS data professionals know by heart). REST endpoints are a nice bonus, but its not enough. It's like asking a chef not to use a knife to cook...