Alle Broadcasts
Fabric Frenzy #1
21 visninger
There is a wealth of new opportunities for data analysis and insight in Microsoft Fabric, and new features are constantly being added making it quite a challenge to stay updated and stay organized.
We want you to be fully updated with the coolest options in Microsoft Fabric. That's why once a month we give you an update on the new features and tips and tricks on how to take full advantage of Microsoft Fabric.
On the last Wednesday of each month, we will make sure to:
- Give you an overview of the most important new features in Microsoft Fabric
- Come up with tips and tricks to make better use of new as well as old features
- Give you concrete examples of cool business applications
View transcript
hi everyone and welcome to the monthly fabric frenzy where we'll go through the news spotlights Spotlight and a bunch of other fun activities so we're going to do this fabric frenzy monthly we are starting today we'll have a bit of a summer summer break next month but we're restarting up the concept in August and then every month we'll be back with any news about fabric any news about power bi and also a bunch of other fun activities for for you to participate in so for today we have an action-packed agenda with the quote of the day but also more importantly Microsoft Fabric and power bi news we have a spotlight today which will be how to organize one leg in the future going forward and finally a chance for you to win home a data model t-shirt so for the data model t-shirt competition we are going to host this on a quick kahoot just to be able to get your information and we'll do a little quick quiz not competing over points or anything but just a little fun quiz and everyone who participates there will be in the draw of of the T-shirt so if you want to prepare and be able to participate in this little fun competition what you can do now is you can go to kahoot by either downloading the kahoot mobile app on your iPhone or on your Android just go to the App Store find it there or if you're on a browser or in a computer or any other device you can do go to kahoot.it and you will be set up there and after you've done that you just need to wait to get the game pin that I will provide you with a later at the end of the session so for the quote of today I watch an interesting video here from SQL Server to fabric by by Patrick and by Bob Ward we're one of the main takeaways here was that but Boyd was able to create a report by loading in the days from SQL then we are also using some c-test on a data Lake and combining them the two in power bi make Dynamic reports and practice ask him the question which is the quote of the day so to get to this point of creating the report how much code did you have to write and Bob of course said zero which is truly a signal of the the power of Microsoft fabric it's also um it's also been reflected on the recent Forester wave that was released in the in the current month where Microsoft is now not just positioned as a strong strategy strategy Legion leader on augmented business intelligence platforms but they're actually almost off the chart it is not possible for them to be any more to the right of this graph they are absolutely the strongest and they are as strong as one can be and if you ask me this is of course to a large extent because of the recent release of fabric but all also all the capabilities that the build up within power bi and with it within Azure synapse over the last few years which is culminating into what we now know and Os as Microsoft fabric so for any anyone out there who are watching this Microsoft fabric frenzy and and don't know what exactly it is still um I'm just going to to do a really quick recap so my Microsoft fabric is a unified platform that brings in all the tools we know and love from Azure synapse from Power bi and from data Factory and adding a few more on the way so what we're getting is that we get the full toolbox of everything we had in synapse we get the tools from data Factory we have everything we have from Power bi and then we get a new tool called Data actuator sometimes sooner and this is all packaged into the platform we already know from Power bi now just extend it with all these new toolboxes and all these new capabilities so very exciting future we have ahead of us with the with Microsoft fabric so what's new in in June for for power bi and fabric well first of all fabric itself is still new but we already covered the basics and there are loads of other things to talk about which is also why we're going to do monthly spotlights spotlights on individual features and how we best work with them and work around them so today most of the news will be about what is new for power bi because there are a bunch of really great really exciting news for power bi that may have been overlooked by the attention that Microsoft fabric has been getting so the first first one I want to bring up is power bi developer mode it's a really cool way to save your work as a new new file format which is is a new symbol very intuitive folder structure that makes it easier to do something like code versioning cicd and stuff like that furthermore we have pricing which was revealed for fabric so we actually know what it's going to cost now to some extent we have new capabilities for embedding power bi it's now easier to work with our power bi outside of the power bi platform and we have a new card visual that really UPS the game of what we can do for making aesthetically pleasing but also very performant very quick visualizations in our reports it's a very very handy new tool for anyone who loves to do neat looking cool reports and finally we have new admin monitoring capabilities that are available to tenant adminants in power bi for a quicker way to get insights that can help you manage administer and govern your tenant and this is all coming out of the box so you just need to go there and click and it'll be ready for you so let's start with the fabric capacity and pricing we've already had the story of fabric is coming with new capacities so where we used to start with what is an f64 was the lowest we had before it was the equivalent of a premium one capacity for power bi we now have lower tiers of capacities going all the way down to an F2 which means we'll be able to start cheaper but also with less capacity so if we are only aiming for the minimum amount of of compute power we could go for an F2 which is currently priced around 263 US dollars per month um as a pay as you go service so that means 263 US dollars per month to get started and you even have to have the capability to turn off the service if you don't need it sometimes during the day or over the weekend and turn it on when you need so basically you it is pay as you go so you pay by the hour Microsoft are already communicating that when we look forward into other billing plans where you can have like a yearly or even more reservations the pricings may be able to be offered even lower than this but just imagine it around 250 US dollars to get all the capabilities of fabric but also all the capabilities of power bi premium at least almost because there's one single feature that is not included it and in anything below the f64 and that is the free Power bi viewer licenses so beforehand and currently if we have an f64 all our reports and everything we create can be consumed and can be viewed by by anyone in our organization without having to buy them a power bi Pro license we only have to buy them for developers but if we're using lower than f64 we have to buy Power bi Pro license to any users who needs to use the the power bi Pro features within that capacity so similar similar to what we're we're used to we can get free licenses with a with a power bi premium one or above and below we have to we have to pay for the pro licenses but we also get cheaper capacity so overall it's a pretty good hybrid approach then we have a new card visual for power bi so beforehand if you wanted to make something like it in the lower left corner here on this report with the with like a card with a with a value and a title and some icon and and a bar to the left we suddenly ended up creating tens or 20 of of unique uh objects and all of these took up some performance and they made the report slow and they made it uh not the nicest developer experience either so with the new card visuals we can actually do a lot of what we had to make a lot of hacks to do we can have multiple cards within one visual we can set up a bespoke designs with icons and sidebars and all these things and it's all just configuration with this in this this Visual and it will allow us to create a very flexible very different set of visualizations that is blazingly fast and performant and some of the common stuff stuff like the example here in the top with headcount turnover rate and cost per higher well that can now be one single visualization instead of being a bunch of different objects so really a superpower for anyone who wants to do neat and great looking reports that are also blazingly fast I'm very excited about that feature then we have admin monitoring which is a much needed feature I mean we can do a lot with the rest API we can get almost all of the data that we need to be able to admin and to govern and to monitor our power bi tenant but it always has been a little bit cumbersome and something that took quite a bit of work with the new admin monitoring we just get a set of data sets and Report out of the box and to begin with we get something called a feature users report and an adoption report and both have its own data set and its own report and they will help us to figure out who's doing what in our organization it will help us to see how usage and adoption is going is it going up is it going down are we paying for licenses that are not used and stuff like that and it can help with Audits and make sure that everything in your tenant is compliant so if you didn't already check this out check if you can access it in your tenant there are still some people who are not able to access the this feature myself included I could open the report but the report had no data so it's still a preview preview thing but but looking at some of the examples of people who are who are using this today it really will once it becomes stable and works for everyone it will be a really really neat addition for any admin of power bi furthermore we also actually got a similar monitoring tool for fabric as a whole which is more of a consumption monitoring but we will cover that in in a in a later session another time then we also have embedding embedding in power bi is sometimes a bit of an underutilized tool because what we can really do with embedding is take anything we create within power bi and make it available for users of other applications so we could put it in web portals we could put it in web applications we could even use it in in notebooks or something like that but it it it needs us to know how to get started and the new big thing is that we got a demo workspace in power bi to help try out the different settings for embedding and it will help us generate the code that we need so we don't have to do so much of ourselves by hand and another feature we got is the ability now to visualize power bi reports directly in jupyter notebooks which means if we're sitting as a data scientist or a data analyst and analyzing some data in our notebooks and we want to do a quick visualization or a quick report just to kind of get an overview of our data we can now do it with a few lines of script inside our notebooks directly in Jupiter notebooks so I think that's a very cool addition then we have the new developer mode for power bi and this is probably the biggest thing for this month if we look at Power bi features because it seems like a small thing we are now able to save our power bi files as a new format it sounds very technical it sounds maybe even boring for something who don't know why this is cool and this is very cool because being able to save our files in this new format allows us to actually understand everything about our data data set and our report inside the file format so we have all the data as metadata in the file instead of being a compressed binary format and that means we can check it by just looking at it we can have third-party applications check it through look at it make sure we follow best practice and stuff like that we can check it into a CI CD setup where we could have code versioning we could have approvals we could have Change Control we can have all these Enterprise level capabilities that will help us make more robust artifacts going forward and all of this can be enabled by a file format that now has and now is I think this is what people refer to when they say desktop pardon that now has all the data and the metadata inside the file format so actually I I installed this on my own personal power bi and I'll show you an example here great so opening up power bi so using this feature works as any other preview features that you may have used already and if you don't know how that works well there is an option inside power bi to enable preview features so I'll quickly go through that go to file don't fill in this one go to file options then options again and there is a tab called preview features and here you can go and enable the power bi project save option down here just click this one and activate it and you will now get this new capability for your desktop as well then if you have a file like I'm opening this pbix because I can also convert a current dot pbix file to the new format so I just opened it here a random Report with a very very small data set of only one table then I'll be able to use save as and now instead of the old way of saving I can change here the type to be a power bi project files now this is actually a folder format so if you just save it in your main folder like download you'll get a bunch of different folders so it's actually important to put this inside a folder so we'll just call this uh fabric demo yeah we'll just call the fabric demo and we'll get the file name here business Central pc280 list which the original report name I save it and that's it I can now use the new file format I can open the report I can edit the report nothing really changed from um from what you call nothing really changed from a developer point of view I still have my report file I can open it and the moment I open it I'm back to using and developing on my on my data set and on my report and while it opened we can maybe explore what's underneath so you will see here so it opens and I can edit just as I did before nothing's nothing changed for the experience of me doing a doing my work but what really changes if we dig into some of these folders because as you can see there's a folder for all what's in the data set and that means all the information on about the model all the information here about everything going on in the model there's a bunch of metadata to dig into and even something like how is the diagram of our of our data model set up so when we actually have these diagram views of our different tables and how they're connected it's actually also defined in here which which allows us to to check it out but it also allows us to change it so we could change some values and it would be reflected inside our power bi same goes for the report we can actually go and check what is going on with the report and some of it seems cumbersome and some of it is even understandable so something like here I can see well this is the main report and here is I have my page one and there's some data data data metadata and there is for example a height value here of the report so yeah if I went crazy and say yeah well I want to make this report Doubles Doubles tall just for the sake of it I could even edit the value here inside the file and this change should be the preview so as they should should now be reflected inside our report we can open it for one last time and check out if it if it really is so and there you go you can see now the report has a has have a different height and that was basically by changing the value inside the the Json files so gives us more flexibility makes it easier for us to rush to Version Control and and things like that going forward and it it really strangens the cable it really enforces enhances the capabilities of what third-party tool providers will be able to let us do with our with our data sets and our reports as objects yeah cool so that was the news for for this month at least the biggest news it's been a it's been an eventful month we got everything in fabric we got also a bunch of other things for power bi but these were in my in my view the most significant and exciting news to talk about um and then we have a spotlight for today on how to organize our one Lakes so one leg is actually at the core either a data leg or a collection of data Lakes um but it has probably it has containers and photo structures as we would be used to it just not calls like that all we can see is one big one leg with a bunch of folders so we have to make some assumptions as of what is going on behind the scenes and we don't really know because that's all up to Microsoft what they're doing but I want to say that we have multiple levels of hierarchies of objects in our one leg so we have the one one leg as a whole things that actually spans across tenants so tenants is not a a top level root level uh object in our one leg one leg actually spans across multiple tenants within those tenants we can have a new thing called domains which is a logical group of several workspaces so you could have something like a finance domain if you wanted to do um do like a business area groups you could have a Germany domain if you're doing geographical grouping or you could even have something like product X if you're doing a data product divisions of your of your domains um all something you need to think about as a company because you can't really mix and match it is it is one of the other and you need to decide which kind of domain structure you need in your organization below the domain we have workspaces that we're already familiar with from Power bi um we can have a bunch of workspaces that can reside in a common domain and and all of them all of the workspaces will actually become a folder in the one leg so from a one leg point of view you have the whole one leg and then you have all the workspaces listed listed at the as the top level folder in the one leg which you can then go into and inside the workspace you can then create artifacts you can create create lake houses you can create warehouses and uh and in those lake houses and warehouses you can store folders and files as well so you have all these options to create hierarchies of of structure so so at the at the gist of it you have here the one leg on the top workspace as the second level of folders then you have domains that are logically boxing in the workspaces but they're not currently being represented in your in your in your actual Lake then you have the artifacts in those workspaces which is also folder levels or folders in the in the one leg and then in those you will automatically get two folders actually three a config folder but also then a folder for your Delta tables and a folder for your custom files and the Delta tables will not be able to fit into subfolders so if you try to do this you cannot do this with the UI but you could do it with notebook or with programming if you try to save a Delta file in the tables folder in some subfolders it will kind of break and not be able to identify itself as a table as a table object so tables are one level without any subfolders but our files we can arrange with hundreds of up for a bunch of subfolders just like we're used to with a regular data lake so we can create any structure we would like inside the files but if we want to utilize the new Delta parquet format inside the tables inside our Lake House of Warehouse we need to keep it as a single single hierarchy which means if we want to structure something like our lake house into a raw layer a standardized layer and a curated layer well then we need to have separate lake houses for those objects if they need to be separated in separate folders because the Delta tables within cannot have subfolders we kind of have a raw folder and a standardized folder and curator folder unless we create a lake house to to host those folders so I'll quickly show you an example of this here is a power bi tenant because it called fabric demo and inside of this I created three lake houses so if you go to something like the Royal lake house you will be able to see that it has the table the table folder and it has the files folder so so um so instead of yeah so so so that's like the the root level of the of the folders but we can also actually dig into the content by using the file explorer that we can install locally so I have a file explorer here for checking out all of my one leg and as you see I have my different workspaces here as uh as the first root level folders then I can go into this and you'll see the different objects here for the raw lake house for the standardized lake house and for the curated lake house so I loaded in some data into the Rolex house and you can see this data is going to tables I loaded them as Delta tables and they will actually be represented as folders themselves so we have a customer with the OK files and the Delta log and everything inside um but I but that that also means that um yeah that everything we do in fabric is actually represented as some sort of folder structure in our one leg and understanding how those are connected can really help by making sure you get the right structure going um going forward cool that was the spotlight or this time um and we'll jump on jump on to the competition about the data model t-shirt so to win this t-shirt you need to I already said this you need to open your kahoot you had some time now to download the app if you wanted to and if you didn't it it actually doesn't take a lot of time you can still do it and you can go to kahoot it and if you're ready to enter the the game pin you will be able to participate in the competition now it's a one question quiz and even just by entering the competition even if you don't answer anything you will still be part of the draw for the price so this is the game pin the game pin is 880.0048 [Music] and you will be able to see it in just a second there you go so you need to go to your kahoot and enter this game pin 880.0048 I've been told that we have a bit of the delay on the stream so I'm surprised I'm excited to see how long it will take for the first people to arrive and there is the first two people I wonder if it's going to be just you two competing here will be interesting to see if there's going to be more but if if not well then it just it just increases your chances of winning so already a good chance 50 50 here I'm gonna give people 30 more seconds to sign up and then we're going to start the quiz so okay okay people are people are entering so welcome to Dennis to tills to style up and to Carlson um into King oh actually maybe I understand underestimated this delay so I should find something to talk about for the next half half minute as well um we can wrap up the one leg bit a bit so we actually have now multi we have we have the workspaces we can use now to create this this structure and we can we can put artifacts inside this workspaces but if you come from a data Lake background you may actually miss out on a lot of the flexibility in making your whole own big scheme of tables if you're wearing a data Lake you can create subfolders of which source does the data come from which type of load are we doing are we doing Delta loads are we doing full loads we can even have versioning on our different loads inside that structure and we can fine tune the performance and everything and this is something where Microsoft is doing this for us and fabric so we don't have the same flexibility but things are more simple and easy so we just need to be aware and comfortable with the new set of design restrictions we are in when we design how we want our one leg to be similar to how we designed our what how we wanted our data Lake to look like before well we have eight consistent eight eight participants so I am a very interested too okay at work quiz so for the fabric events in June 23 this will be the shortest quiz ever done because there's one question and one follow-up question so the first question here is a multi select with no points given so it is just for your sake of seeing how how much you uh you you you got some of these uh these Concepts so in the one leg there are a bunch of objects in in Fabric and there's a bunch of things that creates folders in the one Lake so uh the questions are here you can take like multiple multiple of these will create folders in the one leg will the tenant Creator folder will the domains create a folder well the workspaces create a folder well the artifacts the parquet files or the Delta files that is the question there was a bunch of different answers but the reality is that workspaces artifacts and date Delta files create folders for us so the tenant does not actually for most organizations you will have one leg as the big top tier folder in your in your single tenant so if you're on a single tenant it will feel like the tenant is the top level folder but actually if you had multiple tenants you would you would see all of them as one big one leg so so the tenant is tenant itself is actually not a folder domains is not a folder at all parque files also they are just files but since Delta files is a collection of parquet files with a Json on top it's actually a folder and not not a file the Delta files so um good to see a bunch of people actually got this thing so my my question to you is what are you now most what are you most excited to try out in fabric from what you already know and what you may have seen so while you answer this question I'm going to here do a randomized pick of the of all the contestants in the quiz and after this chance for you to give some feedback we will know who the winner will be and I actually know it now so I just need to stand here in silence uh waiting in anticipation and whoever the winner will be because it will be on the second of when the the session is also over who the winner will be I will announce in the last seconds say goodbye and then I will reach out to to whoever I announce and just I have your email through the kahoot so I will write you and make sure that you get your price accordingly so bunch of answers let's look at them what are people excited to check out developer mode development is also a really really neat feature great so thank you for participating um I draw the winner and the winner is Dennis so congratulations Dennis you are winning a data model t-shirt and for the rest of you thank you for participating we'll be back in August with a new chance to win a t-shirt or maybe something else so thank you for joining hope you have a nice day and probably also hope you have a nice hope you have a nice summer holiday and see you in August thank you