 Yeah, good morning everyone. Good morning, Amy. Good morning, Pierre. You know my coffee Happy Friday. Time zones are hard. We had the MVP Summit community day yesterday and Had we had together a meeting with them Early earlier in the day like at a reasonable time. What was that about one o'clock in the afternoon? Yeah For me it was noon 11 p.m. I Oh Hey, we're global we are global we are global team and As we always say if you need to reach us there's always somebody up It might be Pierre yeah, I top stock at the upstop at Microsoft comm right into our inboxes So Amy any plans on retiring? You know with the latest updates, I think I should I mean I'm getting up there in You know, I have no intention of retiring. I'm on the the pine box program. Oh I'll retire when I'm ready to put be put in the pine box I'm not retiring at all, but something really cool happened this week. I got a One more subscription to a streaming service just so I could catch the alo Premiere on Paramount Plus I Saw the previews for and even my my son who's anti-halo. I know which is weird as a team and I alo Yeah, but he was like whoa. I mean when he's a little kid he dressed up as a halo guy But yeah I need to check it out for sure. I was one of my good friends in Montreal is a Game journalist, so he's a journalist for a tech journalist, but he concentrates a lot on game and game critiques and game development He actually has like Microsoft gave him at the launch of alo. He's got the full-size master chief Statue, you know the one where he's standing on the rock with like a Hero pose he's got one of those in his basement That's cool. I've asked him so many times like how much he wants for it, but he won't sell it That's pretty epic oh Should we get on with the news I guess let's retire some stuff Let's retire some stuff Before we do that, let's look into the chat room we have Andrew McCallum we have Jared Shockley. Good morning gentlemen, and we have some people on Twitch which I can't tell who their names are right now and we are For my first time we're trying it on the LinkedIn live as well, so I'm looking at that one now and there's one attendee, which is me But no for the news first item is If you are managing an API service so API Azure API management zone You really a within the next year, so you've got time One of the reasons I we looked at those items this week and figured okay, let's talk about them is There's a tendency for like when we retired 2008 like for three years. We were telling people 2008 is going please upgrade make sure that It's yours. This is your first warning. This is your 300th warning. This is like we would basically Real now for real now. It's coming and then when we retired it people went what how come you didn't tell us? So this time we figured okay, well, let's talk about them typically we go over some of those that news But we need to talk about these So if you are using Azure API management for your environment, if you're an IT pro and you have to manage that In Order to maintain high level of availability We are changing a little bit the IP addresses of those and we're making them a lot more redundant in seven regions By March 31st 2023 We'll change IP address So that means that your service might be affected So as you see here so in Canada the south Canada central Brazil south Germany west South Africa north Korea central central India and south central US Your service IP address will change There is a process for you to go and Change your system to to the new IP address and your your your IP configuration If you don't do it, here's what's gonna happen your service will still be working So there's not gonna be a downtime the day that it changes. However, you won't be able to Create new API manage the API as you have like you basically won't be able to do much with the APIs that you have so it's a really good idea for you to look at this ahead of time and plan your changes So that's by the time you get to 2023 You're ready and you don't end up with some problems Yeah, it's not broken, but it's not working Well, that thing is we we we tried that the product group they put in these announcements They they but if you're not following their blogs on like an every month or You might actually get to a point where you miss the You miss the announcements So if you miss the announcement and you this is not something you manage every day You've set it up. You've given your developers access to it. They're Deploying their API's to it and your your your Your workloads are actually call actively calling those API's for you It's like your job is pretty much done now. You just have to maintain it But now It are the developers going to come back and tell you oh by the way I saw this and then you have to change all of your IP configuration So it's a good idea to put that on your calendar make a plan and start changing it Change control. Yeah, and all of the details on how to are listed in in the blog post on Itopstock.com I think sometimes you get an email about it too like I remember getting email about deprecated things Yeah, well if you have this service if you have this service, typically you'll get an email, but I know on my end I've had these emails before but I just kind of Not avoided them, but to kind of like skipped over them thinking they were marketing, but they weren't yeah That'll keep me right right The next news item is for you. Yes, we're retiring more things. So again giving that year Warning what's with March 31st 2023? It seems to be like a like a pivotal point in time. Yes Time to retire It's a nexus point as the maybe Marvel universe will Oh So we have some dedicated host SKUs that are retiring some I mean probably very common ones the DS v3 Type one type two So it's obviously it's good news though. You're getting newer hardware with the newer SKU So we just have to you know shut down the deprecated hardware. So you're gonna have to Migrate to the new SKUs before 2023 Cuz they're pretty much saying hey, we're gonna upgrade better get off it or I Don't know. I guess you would be toast. Yeah, I didn't see what the impacts would be if you didn't Well, I think of you if you if you don't migrate your services are just gonna just disappear Gotcha Yeah, not good because that's one thing that's important to understand when we're looking at the way The Azure fabric and the hardware and our data centers are set If it's any machine like regular virtual machine it'll just when we have to replace a rack or replace a pod or replace like the the If you looked at the the the way the generations of of the data center design We went from racks to like containers like the pods and I'm not quite sure I haven't looked in the last couple years As to what which generation we're in but the compute modules or or pods or whatever you will call them They they there's a like like hundreds if not thousands of these compute modules inside this container or pod and if your VM is running there and they're gonna retire this because They're switching hardware the There are the the amount of dead hardware has reached the critical mass within the pod and they have to replace it They don't replace like individual servers. They replace the whole pod. They would just like live migrate Everything off of that to others, right replace the pod and then live migrate it back and you wouldn't even notice Hmm, but when you're in this particular case because they're as you're dedicated to host skews Yeah keyword keyword dedicated host. You're the only person on that host You're the only workload running on that hardware If we have to replace the hardware happens You have sometimes because we're changing CPUs And you know like in VMware like you can't go from like Intel to AMD and back and forth like Easily yeah, there are ways but it's not You basically only get like a subset of the capabilities. I think the same on Yeah, so again Yeah, you have a year to put it on your calendar But yeah, no, I was looking at the the list of new capabilities and it's gonna be valuable for people You mentioned more efficient processors increase availability of CPU vCPUs and RAM Greater regional capacity and Ability to potentially consolidate your workload on a reduced number of hosts or so if the big hosts and Normally, I remember for my customer days usually when we roll out these new skews that we temp people with lower pricing too So it's like hey, this one instead. It's cheaper. So you get the first one for free. All right First servers free Yeah, no, not quite All right, I'm looking at that time we've got lots of time with a go Actually, we're doing really well on time Next item in our list is Template functionality preview in version in Azure portal Will be retired on March 31st 2025 so if you ever say we didn't warn you in time, I Might have to find you and replay this at with you because Like that's three years and it's funny because that template functionality pre you so but never I That's gonna go away, but it's not going away The functionality the way it's working is going away, but now it's being replaced by template specs and table tablet template specs allows you to Easier to manage your templates in the way that you can share them Because I've ran into some problems with with them customers When we're having a template and they put those in their get their their get repository, whichever get service they use and then they're using Different CIC CICD pipelines To deploy their stuff. Maybe they're using PowerShell. They may be the using Azure CL. Excuse me Well, then you have to manage that you start having to manage the certificates for your Your CICD pipeline to actually have access to your private repo Whenever you have private repos and dealing with CICDs and templates arm templates It adds the level of complexity Now with template specs you can actually like basically store them in your subscription and share them with everyone in your subscription so if you decide that the way you set up networking with the the security groups and the firewalls and Like all the restrictions and managements and like network watcher and everything else that you want to because it in your environment This is the way you want to manage things so you can create that one template for vNet and Then you put it in your template specs and then you can share that So if a developer wants to or a developer team wants to set that into Into like a dev environment for For development purposes or testing purposes they can use the same template. So you're really like Cookie cutter those environment, but they're always the same because they all come from the same template So templates that will allow you to do that a lot easier No, that's great. It looks like they added security too. So just looks like they improved the product It's not really it's going away the preview, but here's something better for you. So that's yeah, that sounds awesome because Trying to share was yeah, it was difficult and then yeah, you have your golden standard But someone just spins something new up No Standard anymore wait as a public IP Well, I used to we there was the There was a saying I used to have where somebody would say yeah, but we have build manuals So all our servers should be built the same as it if you give the same build manual to five different system means and say go Build me that server. You're gonna get five slightly different servers Maybe it's gonna be uppercase lowercase Prop the differences which in Windows doesn't really matter but in Linux world matters a lot Maybe it's gonna be where you put the log files versus where you put the like and as build the documentation is never as current as We needed to be right Well, I mean there's fights over naming standards So even in like your VMware environment if you so unused all lowercase, and then so and use all uppercase for the VM But you know, mm-hmm people got Wacked for that. Yeah, so by using templates You avoid this problem because basically you just go take this and stamp it over here I use it and then when I'm done, I delete it, right? So template specs are gonna be a great Product and actually in planning everybody's interested in planning a Azure unblogged episode with the PM for template specs Blueprint and deployment stacks in the very near future Keep watch in this channel. It this is where it'll be Published oh Sorry Jared made me laugh. He said he people always end up logging up the seed Like logging on the C drive. They don't yes They don't add the second drive to your So you end up with a C drive that's completely full and a D drive that's got nothing on and then it crashes But in Azure you have to be worried you have to be careful because the D drive is ephemeral. Oh, yeah So when you're machine shut down that machine that D drive disappears so you can put Temporary logs or temp files on it, but don't put regular logs that you need to keep also for the templates Like I have some templates right now in my Template service in Azure and you can go to the portal and just go with all services and type templates And you'll see if you have any they'll be listed there The product group in the announcement and you can go on the itops talk and see the I put the link to There's a script that the product group has provided That will actually go and grab all of your templates that are saved In your library and actually create template specs Versions of them for you That's cool Be very Be very careful Whenever you download a PowerShell script or any kind of script from the internet You test it first in a sandbox before you run it in production and that goes for even Scripts that you get from Microsoft. That's just that's just Okay Notice the announcement over Warned you I warned you and I'm gonna find you. I have a particular set of skills Okay, next item is yours and it's the final one and yeah, it's a big one. I mean you can help me out, right? Cuz oh And it's not a retirement Yeah, it's actually added functionality, which is really cool So we have the Azure load balancer and now you can make multiple rules on one So you can do the port forwarding and load balancing you can use virtual machine sets or just a group of VMs But what's really cool is you can forward to specific VMs like maybe for troubleshooting or maybe you know I Think the coolest thing I thought With it really helped the scale sets like you're you're starting to scale back down I always wondered like how do people handle that traffic still going and you can just repoint to the other machine in that scale set so It's supposed to help minimize overhead because now you can do multiple rules on the same load balancer So how would that translate in in like day-to-day operation like why would you want to do this? Or what does it bring? Well, I think I mean you have the added functionality So like whether you're using the net that's pointing to load balancer to just point to different machines now You can target specific machines on specific ports as well So as you see the in you know virtual machines in a back-end pool using port 443 And they have different net rules on the same load balancer and then also you can have one for a single machine Which is I mean that's helpful especially Let's say Where would you want it for a single machine? Oh for different ports so you can have the front-end port and And then but yeah the biggest thing I liked was with the scale sets with you know You start scaling up or down you can easily add another rule to include the machine or to Scale down and not have someone accidentally go to a VM that's starting to shut down Yeah, I'm I've read that and after you explained it to me this morning. I went back and checked I find it really interesting the way especially With the new VMSS or virtual machine scale set in flexible mode versus unified Uniform mode new uniform mode you get all your machines are exactly the same and you can't target a specific node of Your of your set You can change something about the set and all of the set changes In flexible mode you can have basically all these machines become Individual machines that are managed as a set, but you can target and manage each of them like with our DP for example individually So having that those multiple load balance rules or not rules inbound that rose For let's say your your workloads running on on the web, so it's port 80 so port 80 will go to the pool over like round robin or number of connections depending what your your your Algorithm is But if you want to RDP into a specific one Because you have that second inbound that rule on a specific port That's a let's say 501 is my first machine 502 is my second machine Whatever the the the the the the that rule you set, but now you can target these individually So if you see in the log for example, you're managing this and you see in the log that one server is always throwing out some exceptions You can actually go into that particular machine from the outside using the load balancer And and check out or even target that specific machine With your workload to see if you can trigger that fault and then do your troubleshooting that way So I thought that was really really interesting opens up a Lot more flexibility in the way we manage and troubleshoot our machines. Yeah, definitely Versus hoping you get you hit the right one, you know Refresh Like a shotgun to the side of a barn and hope that you hit the right board I Remember doing that like hit refraining I keep hitting the one that works. Oh I Thought you meant that thought you meant you remember doing that like shooting at Barnes with the shot. Oh, no No, I can't talk about that Well, you are in the US so that's okay You have a bat I do have a bat and You're not quiet I may pull it out I know and for people that see me every once in a while with the bat in my hands and meetings This is my fidget spinner It's helps me contain my ADHD Which I think on this team. I think we all have it Absolutely Before we jump into the module of the week I just wanted to mention that I believe and Jared on the chat can confirm that that this week is a fourth night and If you don't know what your IT guys are They might because it's a patch and switch week on twitch passion switch our illustrious boss and Yes, he is Is better half? Mr. Joey snow have this wonderful little podcast on geeking out and Just No patch and switch. Oh, okay It is the right week Apparently there is no patch and switch. Well next week is an MVP summit and Well that started yesterday with the yeah pre-day but then next week. Yeah, it is both are busy. Hmm. Oh well We'll just have to figure out something else to do over lunch. We should take it over. I think we could do Playing the role You can do that. Well, I don't have the hair to play the role of Joey Okay, I guess I'll do Joey and then you have the beard for Rick. That's right. And you do have good hair alright, so our learn module of the week is Considering that everything was a retirement except for one So let's focus on the focus on the one so we figured that the learn module of the week with be the Improve availability or application scalability and resiliency using the Azure load balancer It's a pretty good and and pretty quick Learn modules only about 45 minutes ish where you learn how to Identify the features that you can use because learn like that you get the internal low balancer the external low balancer the Like they're the the routing algorithm the the load balancing algorithm There's all of these kinds of things that you need to figure it out Which is not that complicated. We've been doing this for decades Setting up load balancers like hardware load balancers, but this one is virtual in Azure So you go through it to figure out all of the features and capabilities and then you figure out how to deploy it as well, so considering we only had one story about a product that's not retiring Improved it's improved right, but yeah, it's a good intro You know, especially if you're not familiar with load balancing, you know, like the you know, like a five way of GTM and LTM You know, and then this is our version the Azure load balancer Yeah, it's great to learn how to use it Whether it's via external or internal Yep traffic and and Jared in this in the Just wanted to let you know that I wore that shirt for you today and and now there's no show so I'm gonna have to talk to Rick. He's letting me down. I know Sorry man All right, well, thank you everyone and thank you so much Amy for jumping in on the last minute because what we decided Yesterday that you were gonna be my co-pilot. Yeah Then you kind of ghosted me You're doing stuff a lot of stuff. I was doing a lot of stuff plus my daughter showed up right after dinner with a stack of papers and Said can you help me do my taxes? That's such a great way to spend the Thursday night. I'm thinking Picard TV I'm thinking I've got a catch up on the the latest of the the new Halo series and I'm doing You should have a tip today. What's that? Oh? Testing and production, which is Jared and I's a podcast that we have every we alternate weeks typically with a patch and switch And it's just it's our little podcast on Streaming gear presentation skills Tips and tricks on how to set up for like online delivery versus in-person delivery recording Demos versus live demos. We cover like a wide range of subjects and sometimes we get caught on Tensions and we go completely off the rail I gotta check that out. Yeah On twitch in the patch and switch channel Since they're every two weeks, so we basically take the week that they're not there. Okay Yeah, because you are I always go to you for like hey my microphone. Oh god lighting Love I love my new microphone. No, I'm pretty in white and that's lights We need more Rick At the last episode, I'm not sure if it's been put up on YouTube yet, but it will be soon Testing and production All right, so that being said again, thank you very much everyone for joining us Thank you, Amy and we will see you next week on another episode of AZ update Cheers. Have a good weekend. Yep