Evaluating Power BI’s Deployment Pipelines

×

Stephen Wullschleger
Vice President
Senturus, Inc.

Steve’s background in BI, operations and finance spans 30 years. During that time he has led numerous business analytics initiatives using Microsoft data tools and has watched the platform evolve from his initial use of version 1 of SQL Server and Microsoft Access. He brings a deep familiarity with the various capabilities and components of the Microsoft BI platform and understanding of how to best leverage them.

Read more

×

Greetings and welcome to this latest installment of the Senturus Knowledge Series. Today we will be discussing the topic of Evaluating Power BI’s Deployment pipelines.

0:37
Before we get into the main part of the presentation.

0:42
We have some administrative things to cover here. So feel free to use the GoToWebinar control panel to make the session interactive.

0:54
We generally are able to respond to your questions.

2:24
So today’s agenda, we will do some quick introductions, and then we’ll do an introduction to deployment pipelines, including a demonstration, that we’ll have some discussion around requirements, limitations, and recommended best practices. After which we’ll cover a senturus overview, discuss additional resources. And of course, cover the Q&A, which you can enter in the question panel throughout the presentation.

2:55
By means of introductions. Today, I’m pleased to be joined by Stephen Wullschleger, Vice President here at Senturus, with an extensive background and BI, and operations and finance, spanning 30 years. During which time, he has led numerous business analytics initiatives using Microsoft data tools. And as watch the platform evolve from his initial use, a version one of SQL Server, Microsoft Access brings a deep familiarity with the various capabilities and components of the Microsoft BI platform and understanding of how best to leverage that.

3:29
My name is Mike Weinhauer, Director here at Senturus. And among my many responsibilities, I have the pleasure of being the host of our webinar series.

3:39
Before we get into the content, we are going to have, we’re going to do an informal poll, I’ll call it. This is kind of an experiment for us, and that we couldn’t fit into the standard sort of multiple choice format that we normally do. And what we ask you to do is, in the panel, go ahead and, and enter your information, enter it there.

4:06
And the question is, how do you manage your deployment processes now?

4:13
And if you could put that in there, we’ll, we’ll cherry pick a few of those and read it back to you.

4:35
We manage deployment very loosely today. But that’s not an uncommon answer here. And we have, for Power BI reports are deployed ad hoc by the authors. We have more scripted solutions for web apps. OK, that’s it, that’s a great answer.

4:54
Anybody else has any, any responses to that?

4:59
Go ahead and put them in the question panel, and we’ll read them, otherwise, we will move on.

5:07
Steve, I’m going to hand the floor over to you. And go ahead. Thank you.

5:13
Thanks, Mike, and welcome everyone.

5:16
As Mike mentioned, tables, just talk about sort of workflow and kind of managing this whole process of how you develop some reports and kind of do some testing, and then release it out into production. And sort of this new feature that Microsoft launched in, first came in and out in preview back, I believe in May, and scared to go for general lease in this month, or probably next month. And so, it has some features, but also many limitations, which may help with your price process.

5:52
So, it’s something might be, you might be a tool to add to your tool kit. So, the first question is, what is deployment pipelines? And this is a quote from Microsoft. So, you know, they call it efficient and re-usable tool.

6:06
But the main point is, that it’s something for people with premium capacity, and it’s only available if you have premium so on. And it allows you to, basically, step three separate workspaces, one for development, one for test, and one for production. And then it has some sort of some automations to kind of move your datasets, and dashboards, and reports between those three workspaces to progression from dev, test prod.

6:39
So we’ll kind of go through that as an example, and the sort of potential benefits for you is just to help with management of that overall content life cycle.

6:51
And so, finally, they’re starting to add a few sort of work flow helpful features, and then you can work on building a process that’s kind of more efficient and can be re-usable for moving through these three stages.

7:05
And that should provide, for some better productivity, faster delivery of reports, and hopefully, reduced user errors.

7:15
And the best way to get a sense for how it works, is really, I think, through demo. So, I’m here.

7:23
I’m logged into the Power BI service, and for those of you that are familiar to … Service, a key element here is work spaces. So, basically, Workspaces are containers. You can think, like, shared folders on a file server. And that’s where I can have all of my different reports and data sets that. And then, I can grant access to a certain group of users. And everyone can then have access to those sorts of groups of those, those, those reports. And then, you can set up separate Workspaces for different groups if you have different sort of access privileges, things like that. And.

8:04
What you’ll see, if you are a premium user, is something that’s fairly new on the left here. It’s called Deployment Pipelines. So, a new new choice is, if I click on foot plant pipelines, already have one already set up, and I’ll just kinda go to that. so you can give you a sense for sort of the sort of the layout, and then we’ll later, it kinda goes through how to create one from scratch.

8:28
So, as I said, it’s designed to track from the three different States, he does deployment, development, sorry, test and production. And you can see, at each stage, is linked to a specific workspace. So, this case, my development is linked to the sample Dev Workspace. I’ve got tests linked to a different workspace called Test.

8:54
Another one, link to Production. So, each stage corresponds to a Workspace, and I can actually click on the workspace here.

9:01
And it takes and actually launches a new tab, and brings me over that workspace that I was showing you.

9:07
And then, I can click on the Show More, here.

9:10
And you can see all the different databases, datasets, and reports that are in that workspace, and then, similarly, here, I can show the ones there.

9:22
And if you notice, there’s a little difference, because I’ve got, if I’ve got some stuff from the row level security webinar, I did a few months ago, you saw that. And they’re only in my development site, and you can also, then, click on Compare. It’ll actually show you what’s different or what’s changed. And then it’s really kind of is a simple matter than to migrate key.

9:46
Anything you’ve added a Dev or worked on in dev over to test because I can just click Deploy to Test, and it’s going to run for a little bit and just copy those that content across.

10:04
So, give it a second.

10:09
While it’s doing that, there you go, Go to workstation, there it is, OK and I’m gonna actually refresh my browser just to make sure.

10:18
And now, you see, if I show more, you’ll see that there’s the same earned both, and it’s got the little green icon saint area in sync. But now, on the right here, between test and production, this X appeared saying that these two are not in sync. And, one thing to point out, like, if I were a, I could actually, from here. Actually go to the report or the dataset directly.

10:44
So you can see I’ve got this very just simple report, Sayles Byproduct Group here.

10:51
And one thing I’ve done in this workspace is I’ve actually set it up for a setup actually a gateway. Because this is the data source is actually a SQL server running on our local network. So it’s a gateway in order to be able to refresh to the on perm SQL server. And so if I go to that work space here, you can see that I can go to my Dev Workspace and I can just click Refresh and because I’ve got the Gateway setup, it will actually be able to go out and refresh the data. And we can talk more in context.

11:29
If you wanna learn more about the Gateway, a gateway for sort of functionality or setting that sort of schedule goes up.

11:35
But the reason I bring up in this context, if I go back to my pipeline here, notice that in the test section, if I go click on my Report, you say Product Cross. But, oh, it’s going to actually.

11:51
See if this works.

11:56
It’s thinking, it’s thinking, usually, it pops up an error.

12:03
An assignment worked, so there you go. It popped across satisfied, because I did earlier. But the point I’m trying to make is that when it moves the data across from one stage to another, it’s actually not actually moving. The data itself is moving like the schema for the dataset, but not the actual data. She, actually, then, once you’ve moved it across, you have to go and refresh the data. And, most likely, like, from between development and production, you may have different sources, like you want to shift it so that it’s pointing to a different server.

12:40
So, you have to go in and either in the workspaces, then reassign your gateways, you know, kind of go back in and re-establish your sort of your settings here with your gateway connection and make sure it’s pointed to the right one.

12:59
Or within the pipelines if you set everything up correctly, there’s actually an option here.

13:06
Deployment Settings, and you can choose a dataset, and you can assign, actually, I’m gonna go and you can assign rules to it here.

13:18
And that works well. And another sort of recommendation we always make to folks is, whenever you’re connecting to Power BI, and you’re doing your work in Power BI Desktop, is to use parameters for your data sources. So here I am a Power BI Desktop.

13:35
If I click into the Query editor, the way I’ve kind of designed this is, I’ve set, actually set up a source parameter, and I set up a database parameter.

13:46
And then, when I actually queried why I said a query to query my data source, it’s using those per hour parameters and kept to the data source.

13:58
And if I were to actually go, say, click New Source to SQL Server, what I can do is, I can actually choose the parameters as my source, and then it’ll go out to my, my database, and do it that way.

14:13
And we’re gonna actually going to do a whole, sort, of course, or seminar on how to do sort of data refreshes and different ways to set up, collecting between the …, Power BI Service and stuff like that. So this is kind of just a taste of, but the point that you set, these, if you set these parameters, then you do the pipeline. One of your options will be then to set up, to, say, can alter the parameters here. And so, it will automatically replace it, kind of your dev parameter with whatever you want to test. So it’s sort of some built-in sort of replace capabilities. So you can point it somewhere else, but you’re gonna have to play around with it. Whether that’s the easiest way for you to do it, or just going into the workspace, and setting up there, something you have to experiment with, but just be aware of is set up correctly.

15:08
And so then, kind of a Coke gave in, kind of like, the overview, OK? Here’s the pipelines, and stuff like that. So, like, how do I create a new one? Next question. So if I go over to deployment pipelines, I’ve actually can do it from here. I can create pipeline. And I’ll just do that.

15:30
And then it’s going to sell and give it a name, I’ll just give it Type Demo.

15:36
So something and a good practice would be to give it a description, documentation as you go, now is a favorite activity. So, demonstration, demo for webinar.

15:49
And give it a good description for it, with my housemate.

15:52
Come back later and hit Create Ambling to the blank screen.

15:57
The first thing I’m gonna do is assign it to a workspace. So, I assign a workspace to this pipeline.

16:04
It’s going to allow me to select from any Workspaces. I have access to that.

16:10
I have not yet assigned to a pipeline, in this case it’s one called pipe dev.

16:16
And, the point here is, you can only assign one workspace to stage in a pipeline.

16:24
I’m actually gonna assign one workspace.

16:26
Now, it’s going to do, it’s going to allocate that workspace to development.

16:31
And then, I don’t, and then why don’t I, if I click on Deploy To Test, then what it’s going to do, this is actually you can create a workspace called pipe. Give it with this name and then append test after it.

16:48
So if I go back to my workspace is actually new after I refresh the screen.

16:53
Now, you see there’s a new workspace that was created, pipe test, and it’s got those documents in it.

17:00
Then first thing you’re probably going to want to do is think about your naming structure, I didn’t want that little dev and the name, so, and I’m going to take that out.

17:10
And then, you can, I can up here in the screen, view on pipeline in pipeline to take me to the pipeline.

17:18
And now, like I was saying up here, I could work on the parameters, so I can, I can set some data Source rules or some parameter rules, and I can say choose this parameter server, and I can say, give it a different value, for example.

17:37
And then I will create the production version.

17:41
So I can click on Deploy to Production and you’ll see it created a production workspace, and now all of my content is synchronized across all three.

17:57
A few other points to make, there’s the diamond here, the diamond refers to Premium. It means that all of these workspaces are in the premiums, their premium workspaces and the pipeline’s only works with premium.

18:14
And the other thing to point, we ran into, actually, when we’re trying to do a dry run, is you have to really be careful about access privileges. So, the person who’s creating this pipeline has to have permissions to create a premium workspace. Otherwise, it’s not going to be able that are out. There are not gonna be able to create, might create the workspace, but it’s not going to be allocated to the pipeline, because it’s not going to put it in the premium, in your pre instance. So, suddenly aware of. And then, on that topic of access and permissions, there is this, or it’s just like workspaces, you can control access as to who has permissions to create and edit a pipeline.

18:58
So I can add it to the instructor for our classes here, as our instructor, as, if I could.

19:10
Correctly, I can add our instructor as a, as an admin of this pipeline. And then they have the ability to come in and do the refreshes and that sort of thing. One thing I didn’t show you do, they can, can edit the second, refresh the data, where they can migrate things across. As for the, these Dev and test, the test and production workspaces that were created, you’re going to have to go in and manage those permissions, as well, just like any other workspace.

19:44
So, if you look at, So I have to go in here to this workspace, and Go in and edit the permissions there for the workspace, which I would do actually from, from here.

19:59
So I could go to pipe and go to Workspace access, and then I can add whatever members you want to have access to your production workspace, and certainly, most likely, you’re going to create Dev. You’re going to have a set of developers.

20:14
Have permissions there, and maybe a smaller group that does your testing, and then for production, that’s when you mostly have a lot of viewers who are using your kind of end reports. So kind of, man, you can think through then, kind of like the workflow and business process that you want to set up in your organization to kind of manage this process.

20:36
So let me see, just going back to Pipelines here.

20:41
One thing I didn’t mention in my sample is here, all I’ve done so far is when I did publish and deploy, I published all of the documents. But it also has the ability to do selective publishing. So, for example, I did two different rollout, so row level security at attempts, one’s very static and dynamic. And I said, OK, well, the dynamics the one I want to use. And production, And not the other one, So you can actually select the ones you want to deploy to production.

21:18
And it’s not, it didn’t like that, because it also needed the dataset. I was trying to do the report, and also the dataset as well. So now if I publish that across, so you can do that sort of selective deployment.

21:35
So that’s kind of like a summary. So Kyle, like, to review, I can create a pipeline. It will get, gives me, basically, three separate workspaces, and it gives me the ability to migrate the content between the three. And it’s just reports dashboards, and I don’t have a dashboard here, and datasets. It doesn’t do data flows, it doesn’t do some of the other sorts of things you might have. And then you have to be really careful about setting up your permissions and getting the sort of links to your data source.

22:14
Set up correctly so you can do the refresh to show your data.

22:21
So, let’s see.

22:23
Let’s shift back here to the presentation.

22:26
Uh, a few points to make and then we’ll open it up for discussion.

22:32
The first is, as I said, it’s all in premium. So, requires a, an EM or a piece.

22:38
You The second point, just like any other workspace, all, each of each workspace, you have to assign access to the people you want access individually and separately.

22:51
And then, you have to have those permission, obviously, to create those workspaces in order for it to be able to do it, is process. And then, for those of you who’ve been around parvo, awhile, there’s this notion of the new Workspace as compared to the old Workspace and old workspaces were kind of create an audit used to be created automatically. If you create an Office 365 group, and you know, my basically paralyzed and my migraine everyone to try and use a quote new Workspace Shakespeare experience, where I was showing you is really what everyone should be doing going forward.

23:30
And, as I mentioned, permission, this can be tricky, So you want to make sure you get those right, staffing managed separately between pipelines and Workspaces. They’re managed separately.

23:45
And if you have pipeline access, you kind of view the pipeline, that sort of stuff, but not necessarily the workspaces. So you have to have permissions for both. And good practice is as good, just, I know, practice in general. You’ve got your development group, we even without deployment pipelines. On all our projects, we set up a development workspace and a production workspace, and sometimes a test, and separate that work, and just going to setup a workflow.

24:14
More manual deployment pipelines allows it, be slightly more automated. But either way, I mean, as you guys all mentioned in your chat, on the poll, today you’re doing all this manually today. You’re basically opening up Power BI, Desktop, publishing it to one workspace.

24:31
And when you’re ready to publish the next year, we have Power BI desktop, and policy the next. So, this will help a little bit with that process. It’s not like a game changer, but it helps a little bit, automate that process a little bit.

24:44
In that sense, there’s curve.

24:45
You can think of this, sort of use cases, and I showed you kind of like a full deployment, whereas, migrate the whole workspace, or are a selective deployment. And I one thing, when you’re first setting up the pipeline, If you already have, say, in production, a bunch of datasets, and reports, you can use that as the first workspace, and you can, you can assign that to production and it’ll, it’ll create the test and dev and you can actually migrate back and kind of synchronize to go forward. And. Those use cases, where we’ve talked about the sort of, how it’s used, and when it does, but they’re ours is a version one product. There are plenty of limitations.

25:35
First one I’ve mentioned is premium workspaces are required.

25:39
Second thing is all the things that aren’t supported so it doesn’t support incremental refresh. It doesn’t support in some sensitivity labels. It doesn’t support template apps, page paginated reports, or data flows.

25:53
And then the Workspaces themselves are limited to less than 300 items and also downloading the dataset from the target workspace. And so, and then use, as you saw as a thumb rule about the data source rules, they’re actually quite limited in terms of QALY capability, set up those rules.

26:15
And, like said, that the data itself, the contents, didn’t well. So, you can actually publish two apps in a workspace just like anything else, but those app contents and settings don’t get copied across.

26:27
And I think, also, kind of a big deal for folks who are managing big development project projects, assess there’s, there’s no rollback capabilities, There’s no versioning capabilities.

26:41
And there isn’t any sort of workflow built in if you want to say have someone approve, and then promote it to the next stage, and track those approvals and notes about approvals. So, that sort of workflow.

26:56
So, given that, if you’re going to make use of deployment pipelines, a few good practices to follow, treat each workspace as a complete package.

27:05
You think through, you have, you’re going to do this anyway on your Power BI sites. But think through your roles and responsibilities and permissions.

27:17
Think about having an, each data in each stage. Heavier reports, you know, think about different datasets for each stage. And, as I, as I demonstrated, makes, make, make good use of parameters, which is a good practice, in general, from development.

27:33
And an S in development, you’re gonna be using Power BI Desktop to edit it. You can do all your kind of version control, still, outside of Power BI, separate Molly Environment effort there. And then you can use, like, test to check on performance, and get some media usability feedback from folks, and those sorts of things. Then production, you can then, just like you would at managing or production, sort of, workspace.

28:03
With that said, and its limitations, it’s like it has a potentially a role.

28:08
But there’s, just the last point to make is that the next release is currently scale from March, 2020.

28:18
So what you see now is the features are probably not going to change between now and March.

28:24
And there’s a few things on the roadmap, But no, unclear how much they’re going to invest in this, or how much it’s going to grow. Suddenly something to keep sending ideas and request to the community and Microsoft to try and encourage them to do more, if this is something that you want.

28:42
And with that said, it sounds uses, solves all problems for all people with the development cycle.

28:48
Just talk about, you know, other sort of approaches out there.

28:52
So, often if it’s a kind of a large organization, will use kind of git hub as our repository like any sort of software development project for our PTI Axes and then people can check them out and check them in. It’s got, you know, the whole sort of workflow and approval, that sort of stuff, documentation. So you could, you could use Git Hub is kind of your sort of method for managing that sort of process as well. Or, if it’s simpler, if you’re a smaller, you can, you know, SharePoint has some versioning and tracking or OneDrive as well. Seeing how I think about that as another location.

29:32
And I think, Mike, my is when Mike jumps in an after I’m done, it, might be so that you guys have shared it in your chat, which will kinda like bring up as a discussion, because I think part of the point of this webinar is a chance for people to learn from each other.

29:49
So.

29:53
With that, I guess the other point I want to make is that this sort of workflow around deployment is really one component of your overall Power BI Administration.

30:05
So, you think about setting up a senturus of Excellence.

30:08
Setting up your sort of how you’re going to your tenant and premium administration and manage capacity, See, and performance, and optimization, and set your governance and adoption policies. This is a piece of that puzzle. And, so, yeah, I think through your overall sort of Power BI instance, and, and setup, and, and we have a whole course on administration. And if you’re interested with that, I’ll turn it over to Mike.

30:37
Thank you, Steve. That was some great information to appreciate that if you are interested in and things that are related to this topic. And certainly to hang around and get your great questions in the question panel. I’ll go ahead and put your questions. And as Steve mentioned, if you have suggestions, ideas, in terms of the discussion items, go ahead and throw this in there, and we’ll try to bring those up. And Steve, maybe you go peruse some of those questions. We flag there button on our site, senturus.com, on the Resources Site, blog site, We have a white paper that discusses preparing preparations for the Clean Power BI premium as well as a matrix that evaluation looks at different report distribution options by different license types.

31:24
And then we do, actually, it wouldn’t be a senturus webinar without a, an actual poll. So, Steve, I’m going to move to the next slide here. I’m gonna pull up here.

31:35
Whoops.

31:40
Shoot, I closed it by X, say, Yeah, I had opened it. So, did you open it up? Yeah.

31:49
We may have to. I’m not sure we can. Oh, well, I’ll see if I can fix the Paul. While you guys can, again, put responses, I get to this in the chat, if you want, or the questions.

32:03
So given what you know about Power BI Deployment pipelines, how likely are you, are you to use them, so, you know, when you start using them right away, interested, but you’re going to wait for it to sort of evolve, Or you’re not interested in at this time. So you can sort of put that in there. If we manage to reset that thing, then we’ll we’ll come back to it. So a couple of slides about said tourist before we go on to the Q&A against, stick around for that. We have a lot of great questions, and they’re already at senturus. We concentrate our expertise solely on business intelligence with a depth of knowledge across the entire BI stack.

32:40
We are known by our clients, for providing clarity from the chaos of complex business requirements, disparate, data sources, and constantly moving targets, if you want to move to the next slide, Steve.

32:53
We’ve made a name for ourselves because of our strength of bridging the gap between IT and business users, or deliver solutions that give our clients access to reliable analysis ready data across their organizations, enabling them to quickly and easily get answers at the point of impact, in the form of this decision was made, and actions taken.

33:13
On the next slide, our consultants are leading experts in the field of analytics, having years of pragmatic, real-world expertise, and experience advancing the state-of-the-art.

33:22
In fact, we’re so confident in our team and the senturus methodology that we back our projects with industry, unique 100% money back guarantee.

33:32
And we’ve been doing this for a while nearly two decades, across the spectrum, from Fortune 500, have been market companies, soft, like business problems, across, just about every industry, and functional areas, ranging from the office of Finance to sales, and marketing, manufacturing, operations, HR, and IT. We pride ourselves in that our teams are large enough to meet all of your business analytics needs, yet small enough to provide personalized attention.

33:59
The next slide, if you, we have a lot of great resources. We’ve already pointed you over here to send tourist resources for, again, the copies of the webinars and the recordings, but as well as our upcoming events, and blogs and all kinds of great stuff. Speaking of events, we do have some really good Power BI events coming up in the next month. So, month or so, so, go ahead and check back at senturus.com under the events page where you can see all the latest and greatest and register for those upcoming events.

34:31
The next slide, of course, I’d be remiss if I didn’t bring up our excellent training that we are offering in the top three platforms that we support, IBM Cognos. Power BI and Tableau are ideal for organizations that are running multiples of these platforms or those that are moving from one to the other.

34:49
We offer training in all these different modalities, whether it’s tailored group, sessions 1 to 1 or 1 to few mentoring instructor led courses all the way through to self-paced learning and can combine that in any way that suits your organization.

35:06
Then, finally, before we get to the Q and A Some additional resources, head over to senturus.com, where hundreds of free resources, we’ve been committed to sharing our BI expertise now, for over a decade.

35:19
With that, let’s jump over to the Q&A, and I don’t understand what I was able to launch that poll. So, while you’re looking at the question, shall we launch?

35:30
Yeah. Why don’t you want to? Do you want to share it, because I’m not? There we go. It’s open. Great, OK, so go ahead and thank you, Andrea, for resurrecting. Little trigger. Happy on that one.

35:43
So, if you can answer that real quickly, we’ll give you guys a few seconds to do that.

35:54
Caught everybody off guard.

35:56
We go we’re going to get up over half here. Get those votes in here. I’ll give you guys a few more seconds to do that.

36:03
We’re good on time. Seller injury here?

36:09
People aren’t sure.

36:10
OK, we have a little over half. I’m going to close this out and share the results. So, not too surprising, follows that typical product adoption curve.

36:21
20% saying they’re gonna use it right away, and two thirds say they’re gonna hold out for more functionality, or let it mature a little bit, and another 20% that are not interested at this point. So thank you for providing your input there. Actually, Andrea for resurrecting that and interjecting here. So Steve, I don’t know if you had a couple of questions that you guys, cherry picked a few that you really wanted to lead with or if you wanted me to start? I was just going through the ones, the ones you flagged in with a red flag is our two.

36:53
Yeah, that’s exactly. I tag them to sort of say, hey, this is going to ask him. So, I mean, I haven’t called though, I was kind of just going down the order.

37:02
One question was, is a comparison based on the object name or name and version?

37:08
Why? As I mentioned, there isn’t versioning. So it’s, it’s based on the name. So, it doesn’t know enough to is basically based on the name and then I think it’s looking at the most recent dates or something like that or just I’m not sure exactly. How does their algorithm for comparing it? It compares the X similarly between the XML that makes sense. Yeah, mm hmm.

37:32
And then someone’s … was Seen Deployment Pipelines option.

37:39
So that means that, most likely, they don’t have a print it. So the first question was, is it available to non gov tenants? And answer is certainly yes. We’re a non-government tenant.

37:51
He saw that I had it and we’re using Power BI Premium embedded, actually, for this demo. So I’m not sure exactly if you don’t have Premium. That’s the other reason why you wouldn’t see it.

38:04
There are a couple more. You do have to be a pro user. And you do have to be an admin. Newt workspaces as well. That’s right. Yeah.

38:12
Mmm hmm.

38:18
Does the pipeline triplicate the datasets?

38:23
And wouldn’t it cause trouble if you’re close to capacity or premium?

38:25
So, the answer is, yes, It is going to make copies of those datasets. And it’s definitely is going to affect your capacity usage.

38:35
So, definitely, when you’re talking about capacity planning and stuff, that’s going to be something you have to consider for sure.

38:42
Definitely, like a P one node, for example, has 100 terabytes of storage. Typically, the storage is not the capacity.

38:51
Limitation, meaning means.

38:57
Then, another person asked if that you see this approach being used for more self-service, or for enterprise reporting, where more formalized Dev test prod is, is more important.

39:10
Yeah, that’s a good question.

39:15
I see a role for it in both scenarios.

39:20
Yeah, yeah, on the self-service side, kind of if you’ve got like a group of, say, business users that maybe don’t have the sort of no rigid needs, say, of Sarbanes Oxley compliance or something like that, this could be a nice way for them to actually It is pretty simple interface. It might be kind of nice and simple way for them to manage things. That’s not overly complex. And for the larger enterprise, as well, I think about, but they’re also going to have to supplement it. I think, most likely, with some other sort of parts of their process.

39:59
And feel free to jump in others.

40:03
And no data flows is not supported.

40:08
So data flows don’t get promoted as well, they’re not part of that, correct, OK.

40:18
Yes, and so someone’s asking are having different nodes for dev, test, and prod, but our scenario didn’t show that.

40:26
Correct. Our sort of demo environment doesn’t have the multiple nodes.

40:31
I think most organizations would be hard pressed to have that environment set up as well.

40:38
Yeah, larger organizations, obviously, won’t have the funds to do so, but a lot of organizations, which you can, again, you can use those SKUs, and you can ramp up in ASQ just for your dev test promotion, and then cause it to stop funds from stop costs. That is an option we’ve seen with some customers as well.

41:02
Yeah.

41:05
And then, question about data source connections, you can use the same data source connection in all three, kind of workspaces.

41:19
If you’re just one, always connecting the same backend data source, or you can create separate, if you’ve got different data sources, you can create different data, Source connections in each workspace.

41:32
You can have to sift a man, then which one is pointing to.

41:44
Can you remove a report from the data pipeline, and still have in test or production?

41:51
So, the answer is yes.

41:54
So, like, if I, um.

41:59
Go in here right now and say, go to this workspace, my sample Dev, Workspace, and delete this report.

42:13
Then, what you’re going to see in the pipeline here is.

42:22
It’s actually, yes, it did the compare. It shows that there, they’re not in sync.

42:27
So, because that reports that here, but that’s been saying to cross. So, you can have any, but you’ll always get this sort of, you know, compare, you know, icon always being orange.

42:38
If you do that, but you definitely can do it one thing you have to watch out for, the earlier question about, is it report name. It’s actually by the GUI behind the scenes. Yeah. Right.

42:52
If you delete that report from from Dev or let’s say that’s a better scenario if somebody goes or deletes a report from from depth, that’s good. When removes the report from Dev, publishes a new report to Dev, it gets a new good idea at that point.

43:10
So, now, the good idea is between Dev and test and Prod are out of sync. And if you deploy from dev to test, you might have to have the same report, both with different, good ideas in your environments are out of sync.

43:25
So, you do have to be careful with them.

43:27
Yep.

43:31
Let’s see.

43:37
So, how do you create a pipeline for workspace that already have existing reports in production?

43:43
So, if I go here and I create a new pipeline?

43:51
And create it, when I get to this screen, you go to that initial sign-up workspace. Actually, I don’t have any workspaces available, but you could assign that production workspace.

44:04
We, and then it’ll go through and create the other two workspaces.

44:06
It’ll do the backwards, um, process.

44:14
Does contributor role, can they publish between them?

44:19
Um, it’s a good question. You need permissions on the pipelines are different than the permissions on the workspaces.

44:28
So I’ve got the access control on a pipeline, and there’s really all me, an admin option.

44:39
I step in there to visit members?

44:43
Hmm, hmm.

44:45
Members have access to deploy, and Yeah. But contributors do not, by default. There is a setting that you can set at the, I believe, that, the capacity. And so it’s not specific to a single pipeline once you give a contributor access to to deploy it across the full capacity.

45:06
Hmm, hmm.

45:08
Can you bypass the pipeline and publish something directly to Production, not going through the pipeline?

45:16
Yes. Like I could open up Power BI Desktop here. I click on Publish.

45:22
It’s going to give me a list of my workspaces.

45:26
That is it.

45:28
Right. Into that object, ID the good.

45:31
Getting out of sync, exactly. As you are, certainly can, but those, now, all right, replacing that product, do it with a new, and they’re out of sync.

45:42
Yeah, that’s where I was headed with it, as well, as, you really gotta be careful, because then you’re gonna get all the service. He says, things get out of sync, and it doesn’t lose usage miss metrics. You lose those on your old object ID.

45:57
So you gotta be real.

45:58
So if you’re so, if you’re going to go down the route of using deployment pipelines, my recommendation would be that if you’ve got one that you want to push to production, publish it to Dev, and just move through the process to get to production.

46:19
Can’t delete a pipeline if you don’t want to use it. Absolutely.

46:23
So, if I go back here and say, oh, look, I didn’t like this test one delete pipeline, or I can actually, I can actually delete the one from the demo now.

46:35
And then, actually, those workspaces are still there.

46:39
But there are no longer associated the pipeline. You guys actually noticed here. So, if I go to this, this workspace, it’s got the little icon here, which tells me that it’s associated with the development phase of a pipeline.

46:54
And those others, you may recall, I did do that during the demo, but after I deleted the pipeline, they are still here. And all the documents are still here in the state they’re in when you delete the pipeline.

47:06
Um.

47:10
I know Adam, How much is what sort of public about the roadmap? Or what’s known about the roadmap?

47:17
These days, only 2, 2 things mentioned are paginated reports and data flows.

47:23
It’s being supported by the pipeline, mm hmm.

47:29
All right. I think that covered all.

47:31
The one is you marked Mike.

47:35
So, yeah, there are some new ones in there. It’s a flag, so if you see anything in there, we’ve got a few minutes if you want to tackle them.

47:44
Sure, so, no, there.

47:48
Sorry, backward deployment option, for example, test back to dev.

47:56
Um, is there a backward deployment option from test?

48:05
Not really, right? Next?

48:08
At this time, that. Lack of reversions, right, we cannot revert anything at this time.

48:19
Yeah, hopefully, it can.

48:22
Yeah, it’s quite limiting.

48:25
And then, permissions for a pipeline, manage Sheets separately from a workspace, correct?

48:30
So, you can give a department admin, remember X for the dips.

48:34
I’m at pipeline purposes, but keep them at contributor for the Workspace outside deployment pipeline. So, yes, you can do that.

48:46
And if there’s a change in row level security, good question, dev and prod, how do you handle that separately?

49:00
That’s a good question.

49:00
So what it didn’t do, for example, with the row level security, it didn’t move across the, all the everything you would have set up, with, your lowest carry all the security. Let’s see if I go here.

49:19
Um, so, enroll level security, you would have gone to settings, are actually now, you would have gone in, there, a little security. You have gone to set up the security here, and knows, that’s not something I got set up.

49:34
Copied across. So you would still have to go set up your security groups again in the new workspace.

49:42
Does try to segregate those things, so that it is dependent upon your environment. So so if you’re in a test, you may want rather security to be based on a different subset of people, just for you, your UET purposes.

49:57
They don’t deploy some of those metadata items.

50:02
Same thing with refresh schedules, those types of things.

50:17
Well, thank you for, for everyone, for those, those great questions. And thank you very much to Stephen Wullschleger and it’s worth calling out. Mister Adam Harbor, one of the Microsoft experts on our team, who was the other voice. You heard there answering questions.

50:35
So thank you for your great content and thank you all for joining us today. If you want to switch to the last slide there. You see you. Thank you for joining us today. Apologize for the audio challenges, and the pulse snafu there. But we got through it. And we hope to see you soon on one of our upcoming Senturus events. You can always reach us via our website, if you’re old school and actually still pick up a phone, call us at 888-601-6010. Or you can always e-mail us, info@senturus.com for any of your analytics, consulting or training needs.

Read more