As I try to wade through my takeaways from Microsoft Ignite 2017, I first like to think about the most useful sessions I attended there. Let me start by defining “useful” in this context. By “useful” I mean a few things (one, or a combination):
- Any session in which I caught myself saying, “hmm, this is something I feel like I should have known going into this room”
- Any session that got me engaged enough to the point where I was comfortable asking the presenter(s) questions at the end (I think that questions asked during a session is rude, unless the speaker specifically asks for such dialog before starting).
- Any session that had me stoked to get back to work to start testing out in the lab.
- Any session that got me talking to my colleagues about.
Two sessions that met most or all of the criteria above are these (I hope you can view these):
So, the first sessions above sparked some pretty heated debate. In case you didn’t know, Exchange Online runs using the PA, or “Preferred Architecture“. This means that the environment runs on physical servers, with spinning disks (mostly – there are cool new SSD use-cases now, talked about here, also at Ignite, and also an awesome session), aomong many other recommended best practices from decades of Exchange engineering experiences. Check out the volume they deal with – insane!! (*sorry for low quality pic):
One brave panelist in this session let everyone know that she virtualized her Exchange environment and doesn’t care who knows it. Her braveness was rewarded because everyone clapped loudly for her, including me! I don’t discount the value in keeping Exchange workloads physical – I mean, heck, look at the numbers above! However, I think there can be sound business value in virtualizing Exchange when you don’t have incredibly large user bases like Microsoft with Exchange Online. She made it clear that it made sense for her business to go this route. If you’e curious about this panelist, you can find her on Twitter here – she’s awesome. It’s sessions like these that I love because it’s philosophy meets engineering. By this, I mean that from a technical standpoint, it can be done either way. It means that just because you can do something doesn’t necessarily mean you should do something. Its this sort of question that was debated here and it was fantastic. Lots of great input from very smart people.
For the second session mentioned above, this was, ironically enough, the very last session at Ignite 2017. The talk was about how mail flow happens through Office 365, going through the different possible scenarios, like how in a perfect world, your MX points to Exchange Online and all is happy with the world. They talked about how they totally get that this isn’t the case with a lion’s share of Office 365 customers. They know that mail flow can be complicated and convoluted. The good part is that a lot of very welcome changes are coming to help identify mail flow issues, especially around message tracing. There is going to be a LOT more insight into this stuff in the Security & Compliance Center. I, for one, am very much looking forward to this being released. I’ll try to update this post once the session video becomes available because I think every Exchange admin should see what’s coming. It’s incredible stuff. Intro to PP deck below:
I’ll probably post more about other sessions so stay tuned!