Author Archives: sqlswimmer

Using Power BI To Track My Activities

As a MS MVP one of the things you have to do is keep track of all the “things” you do for community, whether it be volunteering, organizing, speaking, etc.  It can be a bit daunting trying to keep track of all of it.  But hey, I’m a Data Platform MVP, how hard can it be to keep track of data?!  Queue music from one of my favorite Blake Edwards movie .. Pink Panther.

At first I was just keeping track of everything in a text file via Notepad.  That got very unmanageable very quickly with all the different kinds of things I was doing.  I migrated all my data to a spreadsheet, because we all know that Excel is the most popular database in the world, right?

I knew that I had been busy in 2018, but I had no idea until I used Power BI to look at my data.  Yes, I was significantly busier in 2018 than I ever had been and 2019 is shaping up to be just the same if not busier.

Take a look at what I created.  It was a fun project to work on and allowed me to explore some things in Power BI that I don’t work with on a regular basis.  Let me know what you think.

Speaking at SQL Saturday Chicago

I am so excited and honored to announce that I have been selected to speak at SQL Saturday Chicago on Saturday, March 23, 2019.  I will be presenting my Profiling Your Data session.

It’s been a few years since I’ve been to Chicago.  Last time I was there was when I was working onsite for a client in Michigan in 2001.  My sister and her son came to visit me over a long weekend and we decided to road trip it to Chicago.  Neither of them had ever been before, so I gave them the grand tour, we road the subway, took in the Navy Pier and even got to see a White Sox game.  To this day, my sister still laughs nervously about her first subway ride.

If you’re in the area stop by and say, “hello”.  I’d love to see you and chat a bit.

Speaking at SQL Saturday Victoria

I am so excited to announce that I have been selected to speak at SQL Saturday Victoria on Saturday, March 16, 2019. I will be presenting my What is Power BI session.

This is another kind of homecoming for me.  When I was a kid, my sister & I lived with my grandparents for a while down in the Willamette Valley and we used to go to Victoria every summer.  I have very fond memories of Butchart Gardens and walking around The Land of the Little People.  I was super bummed when I found out that the latter no longer existed.  But if you have time, I would highly recommend Butchart Gardens and yes, they are open in Winter.

If you’re in the area, stop by and say, “hello”.  I’d love to see you and chat a while.

How To Use Power BI Embedded For Your Customers

Recently I had a need to build a Power BI report for a client.  This client has a multi-tenant database and their own custom web app.  They want to show each client their data without showing any other clients’ data. They also don’t want to require their customers have a Power BI license.  Easy enough, just use Power BI Embedded with the App Owns Data model and set up Roles (row-level security) in the Power BI Desktop file, should take less than five minutes to set up.  Well, we all know it took longer than five minutes or this blog post wouldn’t exist.

There are some great tutorials from Microsoft out there, they even provide a sample app on Github that you can download if you are not a web developer (I am so NOT a web developer!) to help you get started.  There are some great blog posts about it too, one from Reza Rad and one from Kasper de Jonge.  So why I am writing yet another blog post about it?  Because the devil is in the details, and I completely missed the details which means someone else must have as well.

I don’t want to repeat what others have already written, so go ahead, go read their posts, it’s okay I’ll wait here.

Now that you familiar with Row Level Security in Power BI, how do you make it work when you want to pass in your customer’s identifier because your customers don’t have Power BI accounts?  It seems like the only way to make dynamic row level security is to use the Username() DAX function?  But wait, doesn’t that require the user to have a Power BI account?  Sigh, it seems we are going in circles.

The one thing these articles don’t talk about is that when you are using Power BI Embedded, you can pass in whatever you like for the EffectiveIdentity via the Power BI API and it will “overwrite” the Username() function.  What?!  That’s right, it will completely ignore the Username() function and use whatever you give it.  WooHoo!

Let’s see an example, so it’s crystal clear.  I have a small sample of data that has some sales.

Embedded-Sample-Relationships

Let’s say I want to limit the view of the data by Region.  Here are my available regions.

Embedded-Sample-Regions

I would need to create a role based on Region that looks like this

Embedded-Sample-ManageRoles

Notice that I am using the Username() Dax function for Region.  I know, I know, my customers aren’t Power BI users, so this makes no sense, why would I use this?  Like I said earlier, the value that you pass via the Web API as the EffectiveIdentity will overwrite the Username() function.

Now that I have my Roles set in my Power BI Desktop file, I can publish my file to the Power BI service.  All that’s left to do is to add members to my role.  Now when using Power BI Embedded in this scenario, you only need to add one account to the role that was just created.  Just keep in mind that this account must be have a Power BI Pro license.

Navigate to the Security for the dataset.

Embedded-Sample-Navigate-to-Security

Add your user to the role

Embedded-Sample-Add-Account-To-Role

It should look like this now

Embedded-Sample-Final-Role

Now, let’s look at the sample web app and how this will work.  In the homecontroller.cs there is a call to EmbedReport which contains the creation of the EffectiveIdentity, this is where the magic happens.

Embedded-Sample-Effective-Identity

Let’s take a look at how this works from the front end.  Here’s my report with no security.

Embedded-Sample-No-Security

Now I just click the checkbox View as a different user, enter my values then click Reload.

Embedded-Sample-Filtered-Result

Presto change-o, the data is filtered for just the Midwest Region.  But wait, Midwest isn’t a Power BI user, so how does that work?  Again, the EffectiveIdentity overwrites the Username() function and applies whatever you have entered.  It’s magic, I’m telling you.

One thing that tripped me up was that I wanted to filter by RegionID, not region name.  When we work with data, we want to be as precise as possible so we typically use a primary key (or other unique identifier) to filter our data.  However I have discovered through trial and error that you cannot pass a number, you have to pass character data in order for the filter to work.  This means for your data you may need to ensure that there is some unique character data field to identify your customers other than a numeric ID field.

Oh, and if you want to play around with this you can use the sample app from Github and this Power BI Desktop file.

I hope this was helpful.

Update: 2-Feb-2019

I was asked an interesting question via Twitter by Dave Ruijter about converting data to string so you could essentially still use the numeric ID field.

Instead I added a computed column to my dataset that uses the FORMAT() function to convert the existing numeric ID field to a string.

Embedded-Sample-New-Computed-Column

Then I based my role on that column, still using the Username() function.

Embedded-Sample-Using-RegionID-Role

Now I can filter based on the ID field, WooHoo!

Embedded-Sample-Final-Using-RegionID

Thanks Dave Ruijter for making me think about this a little bit more.

Don’t Forget To Take Break

So many times we get sucked into a rabbit whole when fixing/troubleshooting things we forget to take a break.  More often than not this leads to the “not seeing the trees through the forest” syndrome as my grandfather used to call it.

Sadly this happens to me more than I like to admit.  I am tenacious when it comes to problem solving and I don’t like to “give up”.  However, once I hit that mentality of “I’m not going to let this problem beat me” is when I most need to take a break.

I have found that if I quit focusing on the problem at hand and do something else, often times not work related, for a few hours when I come back to the issue I solve it within 10-15 minutes.

Case in point, just yesterday I was troubleshooting a Power BI Data Gateway issue.  We had the Gateway installed and configured.  The Power BI Service could see and communicate with it so we were all set to create our first data source.

We entered in all our information about the data source and when we clicked the Add button, it came back with a failure.  It was a very generic error that didn’t really tell us anything.  I searched the Power BI Community forums only to find that we had done everything correctly.  I was stumped, but I started down the path of “I’m not going to let this problem beat me!”  Right then is when I should have put it down and walked away.  But no, I kept doing the same things over and over again expecting a different result (and we all know what that leads to – CRAZY town).

I was forced to take a break and go coach my swim team of 6-12 year old kids.  I completely forgot about the problem for 2 hours while I coached and talked with my kids.  When I got home and sat down at my desk, I had an epiphany.  Check the Gateway logs!  Wow, why didn’t I think of that before?! I discovered that the gateway had gotten itself into a bad state so I restarted the gateway service and was able to successfully create the data source using the data gateway.  That took less than ten minutes, sheesh.  If only I’d been forced to coach swim team earlier.

So the next time you encounter a problem and just can’t seem to figure it out, take a break.  Give your brain something else to focus on and you just may save yourself from a trip to CRAZY town.

Speaking at DataGrillen

After speaking at SQL Saturday Iceland last year, which is a smaller event that I absolutely loved (you can read about that adventure here), I decided I wanted to do another smaller event.  They are much more intimate and offer a better chance of engaging with attendees, sponsors, organizers and other speakers.

I heard about a smaller event called SQLGrillen that took place in Germany.  Their motto is “Databases, Bratwurst and Beir”.  How cool is that?!  But alas, I missed the call for speakers deadline, so I kept my fingers crossed that the event organizers would put it on again this year and low and behold they did.  They’ve changed the name to DataGrillen, but it’s still the same cool event.

I submitted a session and to my surprise and amazement, I was selected to speak.  I will be giving my Profiling Your Data session on Friday, June 21, 2019.  Now my German is very limited (Nein, Ja, Bitte, Bier & Bahnhof) but I’m still excited and I hope to see you there.

The event is already sold out, but they do have a wait list.  If you want to go get your name on the list as soon as possible because it’s FIFO.

Auf wiedersehen for now.

Speaking at SQLBits

I got an early Christmas present this year, I found out I had been selected to speak at SQLBits!  That’s what I call a gift that keeps on giving.

I have always wanted to attend SQLBits so I decided that 2019 would be the year I would finally attend.  Since I had decided to attend, I thought, “what the heck, why not submit session?  I’m going to be there anyway.”  But never in my wildest dreams did I ever expect to be selected.  I will be presenting my Profiling Your Data session.

It’s been twenty years since I was in England and I am super excited to be going back.  I have family ties to England so I added a few extra days for sight seeing.  Last time I was there I visited Malvern Link, home of the Morgan Motor Car Company, my dad’s favorite auto manufacturer.   This time I am planning a quick trip over to Liverpool so I can see where my Dad’s favorite band (and one of mine) was started, you may have heard of them, they’re called The Beatles Winking smile