Month: May 2018

Yes, another post with an obscure reference for a title.

After some time discussing tech subjects, I was of a mind of going back to something that has often been misread in the past by IT teams and IT management. And by that I mean : business. Yes, again.

Do not misunderstand me, I am still a technologist, and I love learning about technology, finding out the limits and possibilities of any enw tech that is coming out. I am not a sales person, nor a marketing person. However I have been exposed to many well crafted presentations and talks over the years, and what often came out of even the most interesting ones was that : “our tech is fantastic, buy it!”

All right, I love that tech. Be it virtualisation, SAN, VSAN, public cloud, containers, CI/CD, DevOps… choose whatever you like. But technology is not an end to itself in our day to day world. Whatever matters is what you will do with it for your company or customers.

I will take an example. An easy shot at someone I admire. Mark Russinovich, CTO of Azure, and longtime Windows expert (I would use a stronger term if I knew one 🙂 ). A few months ago, during a conference, he had a demo running where he could spin up thousands of container instances in a few seconds, with a simple command.

First reaction : “Wow!”

Second reaction : “Wooooooowwww!”

Third reaction : “How can we do the same?”

Fourth reaction (probably the sanest one) : “Wait, what’s the point?”

And there we go. What was the point. For me, Mark’s point was to show how good Azure tech is. Which is his job, and this demo made that very clear. But Mark did go further, as he usually does, during his speech and encouraged everyone to think about the usages. Unfortunately, most of the people I have discussed with seem to miss the point. They see the Wow effect, and want to share it. But few of us decide to sit down and think about what the use case could be.

And that is the difficult, and probably multi-million dollar question : how to turn amazing technology into a business benefit.

Never forget that, apart from some very lucky people, we are part of a company that is trying to make money, and our role is to participate to that goal. We should always think about our customers, internal or external, and how we can help them. If doing that involves playing with some cool toys and be able to brag about it, go for it! But that is not the other way around.

PS : to give one answer to how we could use Azure Container instances for the real world, especially the kubelet version of ACI, try and think about batch computing, where you would periodically need to spin up dozens or hundreds of container instances for a very short time. Does that ring any bell for you?

PPS : I could not find the exact session from Mark I am describing here, but there is an almost identical session from Corey Sander and Rick Claus there : Azure Container Instances: Get containers up and running in seconds

There we are, I have finally given my talk about Kubernetes and Azure.

It was both more and less than I expected.

It was more easy, once I got there, into the position of a speaker than I expected. My fellow speakers were very kind and supportive, which helped with the pre-stage flutters 🙂 It was also easier because the room was of a reasonable size, and I was not on stage in front of 500 people.

And it was less deep dive than I expected, which also allowed me to relax a bit. I could get a feeling about the audience before going there, which let me into the dark regarding their needs and expectations.

 

Let’s set the stage. The event took place at Microsoft’s Building 20, which is a Reactor (https://developer.microsoft.com/en-us/reactor/). So the building is definitely designed to host events comfortably. That helped a lot, as we even had someone from the A/V team to help us and ensure all the screens and microphones would be working correctly. And yes, the free coffee might also have been a huge help 🙂

The room was large, without any raised platform for the speaker, but with multiple repeat screens all around.

I was the third speaker, so I definitely had some time to review my slides and demo setup a few times.

I did setup the demo environment the night before, to avoid any deployment issue at the last minute (which did happen 2 days before while I was practicing). Once again, having a scripted demo ensured that I would not forget any step, or mess up some command line options.

 

I did have a few issues during the talk. First the mike did stop working at some point, failed battery. I kept on speaking without it, as the room was small enough to let me speak louder for a short time and still be heard. The support guy came shortly to replace the battery, so no big issue there.

My remote clicker did work perfectly, but not the pointer part. That’s a shame, because it made it more difficult to point out at a precise section of a slide or demo. Afterwards I found out why, and I should be able to avoid that particular issue in the future.

I did not get as much interaction as I hoped I would. I thing that it was mostly due to my anxiety, which prevented me to behave like my normal self and be engaging.

 

What I would change for the future. First, for a set event like this one, I would practice in front of a camera, or a mirror, to actually see and listen to my speech. That would probably ensure that I would keep the correct pace and articulation. And also make sure that the flow of slides is comprehensible.

Second, I would work more to know the expectations of the public. It turns out that my talk was way too technical and fast it should have been. While discussing with the attendees afterwards, I realized that I did not get many of the points through, probably because I went too fast over those. This brings me back to the interactions point above : would I have been more comfortable and interactive, I could have grasped that during the session and corrected it.

Third, I should probably think about learning a bit more about controlling my voice and projecting it. I realized that during the week leading to the event, as I had to speak in a loud environment, and present/discuss the same kind of subjects.

 

Labs

A word on the hands-on labs we had in the afternoon. I just was glad to have stayed for that part.

First because I had never been on the proctor side before, and it’s really fascinating to see a problem through the eye of someone with a different mindset and culture. I really learned a lot, and realized a lot during these 2 hours.

Second, because it showed me the areas where my presentation had been lacking, and how much I had not been clear enough to be understood by everyone.  I think these discussions with the attendees were the deeper feedback and improvement tips that I could get.

For the record, the container labs we used are there : https://github.com/Azure/blackbelt-aks-hackfest/

That’s it for now. This first talk has unlocked something and made me realize that I should talk at every occasion I can, and that I love it, at least when it’s done 😉