Highlights from our Automated Testing in Qlik live event
We have officially hosted our first live event and thought it only appropriate to gather some of the best material and produce a written highlight reel. To watch the full chat between our founder Oskar Gröndahl and Carl Andersson, our Head of Sales, all you need to do is click play – they discuss automated testing in Qlik and the birth of NodeGraph. Otherwise, enjoy some of our favorite tidbits below.
NodeGraph’s origin story – filling the gaps in Qlik
“When we started off working with BI and Qlik like 10 years ago (10 plus years ago, actually), it was very different to what BI, and Qlik in particular, is today. Back then, when we did sales with Qlik, the typical customer would be [working] directly [on] the business[-side], so maybe a CFO or controller. And, very often, the IT department was not even part of the decision of buying Qlik. And what that means is when it comes to, you know, uptime of Qlik or when it comes to the data quality of your Qlik solution, it wasn’t really that important.
So, you know, when the CFO called me like, “Hey, you know our sales dashboard is not working today”, it was more like a high-five situation – it was not a big deal. We hooked on to remote desktop, we fixed the dashboard and everything was all good. But [in] the last 10 years, we’ve seen a shift when it comes to the Qlik product, right? The product itself has become a more enterprise-ready product, now have Qlik Sense as well, we have the NPrinting solution as part of the Qlik package. And what this means is that the Qlik product has moved away from the business-side and maybe moved a bit more towards IT. And, also the user base within the customer has gone from maybe 5 or 10 users within their company to maybe 500 or 5,000 users. So, of course, the demand of running a solution like this is way different than what it used to be in the beginning.
So, what we saw, and I think it started maybe 5 or 6 years ago, is that there were a couple of gaps as we see it within the [Qlik] product. From the sort of development standpoint or the back end when it comes to data lineage, i.e. where your data comes from, it was very difficult to find that within the Qlik solution. So we saw a space there and that’s one of the reasons why we started NodeGraph.”
From a tool to a platform
“When we started NodeGraph, we were focusing on the lineage part and automated documentation, of course. That’s a huge benefit with NodeGraph – you don’t have to manually update your documentation which is, I mean, for anyone out there doing development within Qlik, that’s a huge benefit, right? But we started off with lineage, automated documentation and we then moved on to the data governance part where you can actually get a very good understanding of your Qlik solution and what’s going on within the solution. And now, since we added our last module which is focusing on data quality when it comes to regression testing and unit testing, yeah now I would say that we’re more of a platform and we’re working with platform initiatives. I think it’s very interesting.”
What is unit and regression testing?
“Regression testing is making sure that what used to work is still working within your Qlik solution. For instance, you know the data in your solution should not change because you make minor changes with new scripts. So, if the sales amount used to be 50 million for 2015, it should still be 50 million after this new release.
That’s the whole thing about regression testing and, traditionally within Qlik, we haven’t really been speaking about testing that much. But now, I mean testing is a vital part for every tech product out there and this is what we would like to bring to the table.”
Testing in action
“Say, for instance, we have a problem that the data is not being updated [and] we have this weekly report coming out Monday morning – it’s super important for our controllers around the world that the data is up-to-date. So, before you had these test possibilities within NodeGraph you’d probably have a person actually waking up early Monday morning making sure that the day is up-to-date manually.
So now what we do is that we have a test case for this – we tell NodeGraph every Monday morning 4:00 a.m. test this data. If it fails, NodeGraph can create actions to post to a Slack channel [or] email or whatever notification system you like.”