Shadowing Practice: Azure App Service and Virtual Network Integration Options - Learn English Speaking with Video

Hey everyone.

In this video, I want to talk about the relationship and the interaction between app services and virtual networks.

Because there seems to be a lot of questions coming in about,

well, do I use a private endpoint?

Do I use VNet integration?

Do I use a gateway?

So let's kind of look at this.

So app services were actually one of the original Azure services way back

when it first started and gone through a lot of innovation around now.

So I can think about,

well, I have my app service plan.

Now, normally this is a multi-tenant model.

There's a certain stamp of infrastructure that has front ends.

There's data roles.

There's file services.

So that is shared by the customers on that particular staff.

There's a single inbound IP address.

You can get a unique IP address if you really want it through the kind of SSL option.

Then there's multiple outbound IP addresses.

We have this shared set of front ends,

the data roles, the file servers.

And then unique per customer,

you have a number of workers.

These are the things where you say,

hey, this is the type of worker on the SKU.

These can auto scale.

And this is where when you create your particular kind of web app,

etc., it's running on your workers.

So the workers, they're unique to you.

Then there's that other shared set of infrastructure.

And then you have kind of your virtual network.

So over here I'm creating my VNet,

which again is kind of a set of IP addresses.

And the challenge we have is really two.

The first one is the apps running in my app service plan

want to be able to reach resources running in my VNet.

And or, well maybe there's other VNets that I have tiered them.

And or there's on-premises resources that I happen to have connected via site-to-site VPN or ExpressRoute.

So from the app, maybe there's a database,

there's some other tier of service.

I want to be able to get to things here.

And then I've got resources running here that want to be able to privately get to the app.

Remember, the multi-tenant model, there's an inbound IP address that's public facing.

And maybe I want to lock that down.

So what can I do with that?

So we have these two different requirements.

Now, when we think about that,

let's start off thinking about going to our application.

And the first solution is kind of a very familiar one.

It's service endpoints.

So if we think about,

well, hey, look, we've got our virtual network here.

And we think about, well,

we divide this into subnets.

So I can think about,

well, I have a particular subnet here.

And what I can do is I can turn on the service endpoint.

So the service endpoint is going to be that Microsoft.web.

Now, this particular subnet can be known to app services.

So now on this app service,

it has kind of its set of inbound controls.

I can actually specify, hey,

well, on those inbound controls,

you can imagine it's kind of like a firewall into the app service.

I'm only going to allow in,

and we'll call this subnet one,

well I'm going to say subnet one, yes, you're allowed.

So it's still technically going through the public IP address,

but when I do the service endpoint,

remember it creates the optimized route,

so it's not just bouncing around the regular edge routers,

etc. It's doing a very direct route,

really as optimized as it can get.

And I'm now restricting it through its rules to say,

hey, only coming from this particular subnet.

Now the challenge with this is service endpoints,

it's just for things in the subnet.

What about if I want to get to it from other networks or on-premises?

So one of the things I can kind of add to this is,

well, I can actually do something like AppGateway.

If I deployed AppGateway into that subnet with the service endpoint of Microsoft.web,

other things would come into the AppGateway and essentially proxy through and then go via that path.

So this gives me that ability.

So it's still going to the public endpoint,

but it's completely locked down.

I can't get to it from anywhere other than the subnet to the service endpoint that I have enabled.

So it's all about, hey,

from our vNet, wanting to go to that particular app.

The other option, well, there is one other,

but the other main option for controlling giving access that way is private endpoints.

So that's obviously very, very common today.

More and more services are kind of adopting this.

So the private endpoint, there's an IP address is consumed from a particular subnet in our virtual network that essentially

is natting the traffic to that.

I could now completely again lock down the app to not allow anything other than this particular private endpoint.

And that's going to work for kind of Windows apps,

Linux apps, even Windows containers today.

So at this precise moment,

as I'm recording, it's previewed,

100

but I think it's going to come out of that any day.

101

So again, that's the other option,

102

private endpoint, I would then access that IP address.

103

And the great thing about private endpoint is it is just an IP in this VNet.

104

So any network that is connected to this VNet can see the IP and use it.

105

I just have to make sure I have the consistent DNS in place so it works and resolves to the private endpoint.

106

I mentioned there was another.

107

There is, of course, IP address restrictions that I can use on this thing.

108

So if I know the IP address the request is coming from,

109

I can restrict it to just that IP address.

110

So if, for example, I could have a NAT gateway,

111

I could have a standard load balancer with outbound rules,

112

I know the outbound IP,

113

I could use that as well.

114

But honestly, if I'm talking about something in the VNet talking to it,

115

that service endpoint is a better option.

116

It's really locking it down to that particular subnet. So great.

117

that addressed things going from the VNet talking to the app.

118

What about the other direction?

119

Now, my application wants to be able to talk to things actually within the virtual network.

120

Now, option one is not really well suited to this.

121

Option one is suited to the idea that,

122

hey, there's some resource on-premises,

123

could be a database, some other component,

124

that I want the app to talk to.

125

And this is where we actually go and use hybrid connections.

126

So with hybrid connections, we have this hybrid connection manager that we deploy.

127

It establishes an outbound connection over 443 to Azure Relay,

128

which then lets the app talk to the service.

129

Now, this is only TCP.

130

It's a particular TCP port and endpoint.

131

So I'd have to have one of these for each different sort of sets of things I want to talk to.

132

It's TCP only.

133

I can't do UDP over this thing.

134

But now it's established the outbound connection which now enables the app to go

135

and talk to whatever source this is kind of offering.

136

It could be a database, for example.

137

So technically, I could deploy a hybrid connection manager in my VNet.

138

It would establish that outbound 443 to the Azure Relay,

139

which would now enable that direction of connectivity.

140

So absolutely, I can do that.

141

It doesn't make the most sense in the world, though.

142

There's better solutions for that. And we have two.

143

So the first is a gateway required VNet integration.

144

So as the name kind of suggests,

145

I have to have a gateway.

146

So I can think about it,

147

we use a different color.

148

So here I'm gonna have a gateway.

149

And it has to support point to site VPN.

150

So we're gonna do the route based dynamic.

151

And what's gonna happen here is my app is essentially gonna establish a point to site VPN connection.

152

So it is gonna go and connect to the gateway,

153

so I have to have the gateway.

154

Remember the gateway lives in its own kind of dedicated subnet.

155

And then from there, well,

156

it would be able to talk to those things.

157

Now, this does not allow me to go and traverse the things like ExpressRoute.

158

What is nice about this gateway option is,

159

well, actually that gateway where you could live in any Azure region.

160

So maybe the app service plan is South Central.

161

I wanna go and talk to a VNet that's in East or West or Europe.

162

This will work.

163

I can have a gateway in other regions and it will be able to go and talk to it.

164

Also, this will work with classic virtual networks.

165

So the old style before ARM,

166

I could have the gateway there and I could do a point to site VPN connection to it.

167

So that's one option.

168

The preferred option is regional vNet integration.

169

So we're going to kind of draw this in yet another color.

170

We'll do this in gold.

171

So with regional vNet integration,

172

as the name kind of suggests,

173

we have a particular subnet.

174

Now this is going to be delegated for this app to integrate in.

175

So now what we're going to do is this app is essentially going to take over this subnet,

176

and it's going to consume IP addresses within that subnet.

177

This has to be in the same region as the app.

178

That's why it's regional VNet integration.

179

I cannot use this if I want to talk to a VNet that's in a different region.

180

I cannot use this to talk to a classic virtual network.

181

So this is the best solution if the VNet is in the same region as the app.

182

But if it's a different region,

183

it's classic, I'll have to go for the gateway approach.

184

Now, what's going to happen here,

185

remember these workers, each worker is going to consume an IP address in this kind of delegated subnet.

186

And I can't use this subnet for anything else.

187

it's locked down only the app service plan.

188

So if I had three workers,

189

I'd be consuming three IP addresses.

190

So when I'm sizing this subnet,

191

I have to think about what's the maximum scale I'm ever gonna do,

192

and then double it.

193

Because let's say I was gonna have eight workers at max.

194

Remember, if I resize my workers,

195

the way Azure works is it spins up eight new ones of the new size,

196

make sure they're working, and then deletes the old one.

197

So it'd have the eight existing,

198

the eight new ones, and then it would delete the old.

199

So I'd have double that number.

200

So when I think about sizing,

201

make sure it's double the maximum number of workers you're ever gonna have.

202

Remember, you lose five IP addresses.

203

So the host, the broadcast,

204

the three Azure ones that it steals.

205

So make sure you size this subnet.

206

So you're gonna size this subnet.

207

So if I thought I was gonna have eight workers,

208

I think, well, I need 16 usable IP addresses,

209

then I'd probably make that kind of a slash 27.

210

If you're not short on IP addresses,

211

maybe just give it a slash 24.

212

But make sure you size it correctly.

213

So from here, it can now go and talk to things in that virtual network.

214

It can talk to things over the express route.

215

So I can go and talk to that as well.

216

It cannot talk to peer networks.

217

So if I had another network that's peered, it's not gonna work.

218

If it was a regional peer, it's not gonna work.

219

That's what's called regional Vignette Peering.

220

It doesn't support peering itself today.

221

It's just a regional integration.

222

Now, if I had private endpoints within this virtual network talking to other services,

223

maybe I've got a particular storage account or something.

224

Well now, this app through the Vignette integration can go and talk via the private endpoint to that storage.

225

So I can start to lock those things that way,

226

it can work together like that.

227

So what do we kind of have to the app?

228

I can do IP access restrictions,

229

but generally we're gonna do a service endpoint for the Microsoft.web.

230

If I need beyond the subnet,

231

I can always proxy via App Gateway,

232

or I can use private endpoints.

233

from the app to my resources,

234

well, yeah, I can use the hybrid connections.

235

Doesn't make the most sense.

236

I can use the gateway,

237

point to site, works across different regions.

238

Best option is the VNet integration.

239

Today, VNet has to be the same region,

240

actually has the app itself.

241

One caveat to all of these things,

242

and there's different types of kind of app,

243

like running an app service plan,

244

web apps, mobile apps, API apps, and functions.

245

Now remember, functions can be serverless.

246

If I wanna use functions,

247

and I wanna use these kind of capabilities,

248

with the exception of the service endpoints and the IP restrictions,

249

if I wanna do functions,

250

I have to run it in kind of a dedicated,

251

I am running it in a regular app service plan of workers,

252

or I think it's kind of the elastic premium,

253

which is fairly close to a kind of dedicated.

254

I cannot do the pure consumption,

255

the regular consumption, can't spell,

256

functions will not work for most of these things,

257

because it can't, there's no dedicated set of resources establish this.

258

So if you want to use Azure Functions,

259

I want to use private endpoints,

260

I want to use the VNet integration,

261

I need to run it on a dedicated App Service Plan or Elastic Premium,

262

then I can get that.

263

Now there is of course one other option and this is the ACE.

264

So if I just draw a brand new picture for this one super quick,

265

I can think about once again,

266

I have my virtual network,

267

so I have my VNet.

268

And remember before we talked about there's all those shared components like the data stores,

269

the file servers, the front ends,

270

and then there was the workers.

271

With an ACE, what actually happens is I have a particular subnet,

272

and then I deploy my app service environment into that directly.

273

So I can think about within here,

274

I have kind of the front ends,

275

I have kind of the file servers,

276

and I have my workers.

277

So these are now actually running in the subnet in my virtual network.

278

It's dedicated, it's all dedicated to me.

279

So now if there are other resources in the vnet

280

or cover connected well there's no other integration required this stuff's

281

sitting in the subnet there's different types of ace there's internal

282

external in terms of what it's facing you would do internal

283

if you wanted that direction but now it's on the subnet

284

inside that ace i would create one or more app service plans where i'd run my apps

285

and they'll now have that full kind of connectivity to it.

286

So that's the other option.

287

The downside is the ACE is more expensive because I'm not getting the benefit of those shared components anymore.

288

I kind of split the cost with other tenants and they're all running inside my subnet.

289

But this was kind of the traditional way we had to have kind of that private connectivity in the past.

290

But now we have all those other great private endpoints,

291

service endpoints, Bnet integration, etc. One note,

292

if you do this, remember there is still kind of the Azure Resource Manager management plane and lots of other things.

293

This still has to be able to talk to that.

294

If you do a bunch of controls and locks and things on this and lock it down so much,

295

you can break it.

296

There's actually a lot of different communications required.

297

They're all documented, but don't think I can just deploy an ACE and then turn everything off,

298

it's gonna break.

299

So there's still communications required with the kind of arm management,

300

but this is another option.

301

It's just they are in your vNet.

302

That wouldn't be my first choice because of the cost.

303

There's more things involved.

304

Ideally, you'll kind of look at these And in the ideal world,

305

it's the same region.

306

I'm probably going to use the regional network integration option and private endpoints or those kind of service endpoints.

307

We'll hopefully do what you need.

308

So, again, I hope this was useful.

309

I was trying to clear up the distinction.

310

I'm going to need one of each.

311

VNet integration doesn't provide me with the ability to talk to the service.

312

VNet integration is app talking stuff here and then private endpoints or service endpoints go in the other way.

313

So, it's different technologies.

314

they're all unidirectional they do one of the requirements we have hope this was useful again please like comment subscribe

315

and share if it was until next time take care

✨ Aanbevolen Video

497 zinnen

Amazon Bedrock Tutorial | Getting Started with Amazon Bedrock | AI Model Access & Testing

Business EnglishUniversity StudentsWorking Adults

257 zinnen

Learn Next.js Parallel Routes In 16 Minutes

Business EnglishUniversity StudentsWorking Adults

Woordenschat en spreektips bij deze les

Deze spreekles op niveau C1 is gebaseerd op de video “Azure App Service and Virtual Network Integration Options”. Deze woorden komen het vaakst terug: app, endpoint, subnet, vnet, network. Deze video bevat 315 zinnen en 2881 woorden om na te spreken. Het gesproken deel duurt 19:41. De spreker praat in een gelijkmatig tempo van ongeveer 146 woorden per minuut, prettig om te shadowen. 82% van de woorden hoort bij de 3.000 meest gebruikte Engelse woorden; de rest kun je beter vooraf bekijken.

Belangrijke woorden in deze video

De 15 moeilijkste woorden uit de video, met uitspraak en betekenis:

Woord	Uitspraak	Betekenis
dedicate werkwoord	/ˈdɛdɪˌkeɪt/	inwijden, wijden
deploy werkwoord	/dɪˈplɔɪ/	uitrollen
consume werkwoord	/kənˈsum/	consumeren, verbruiken
restriction zelfstandig naamwoord	/ɹɪˈstɹɪkʃən/	beperking
tenant zelfstandig naamwoord	/ˈtɛ.nənt/	bewoner, huurder
premise zelfstandig naamwoord	/ˈpɹɛm.ɪs/	voorwaarde
optimize werkwoord	/ˈɑptɪmaɪz/	optimistisch zijn/handelen
elastic bijvoeglijk naamwoord	/ə-/	elastieken
delegate zelfstandig naamwoord	/ˈdɛlɪɡət/	afgevaardigde, gedelegeerde
proxy zelfstandig naamwoord	/ˈpɹɑk.si/	procuratiehouder, gevolmachtigde
restrict werkwoord	/ɹɪˈstɹɪkt/	beperken, begrenzen
caveat zelfstandig naamwoord	/ˈkæviɑt/	waarschuwing
firewall zelfstandig naamwoord	/ˈfaɪ(ə)ɹˌwɔl/	brandmuur
resize werkwoord		afmeting veranderen
router zelfstandig naamwoord	/ˈɹaʊ.tɚ/	router

Phrasal verbs die je zult horen

Woord	Betekenis
take over werkwoord	overnemen

Zinnen om te herhalen

Korte, volledige zinnen uit de video die je in alledaagse gesprekken kunt gebruiken:

Do I use a gateway?
We'll do this in gold.
It's dedicated, it's all dedicated to me.
It's just they are in your vNet.
We'll hopefully do what you need.

Grammatica in deze video

De structuren die de spreker het meest gebruikt, met de exacte woorden uit de video:

Structuur	In de video
Lijdende vorm be + voltooid deelwoord — het gaat om wat er gebeurt, niet om wie het doet	is shared · can be known · is consumed
Present perfect have/has + voltooid deelwoord — iets uit het verleden dat nu nog telt	have tiered · have enabled

Uitspraak om op te letten

De spreker gebruikt 69 samentrekkingen en verkorte vormen, zoals gonna, I'm, can't. Spreek ze kort uit, zoals je ze hoort.

De klanken “sh” en “zh”: restriction /ɹɪˈstɹɪkʃən/, unidirectional /ˌjuː.nɪ.dɪˈɹɛk.ʃə.nəl
Lange woorden — let op de klemtoon: unidirectional /ˌjuː.nɪ.dɪˈɹɛk.ʃə.nəl, ideally /aɪˈdi.ə.li/, automobile /ˈɔː.tə.məˌbil/, capability /ˌkeɪ.pəˈbɪl.ɪ.ti/

Klanken die Nederlandstaligen lastig vinden:

Stemhebbende eindmedeklinker — /b/, /d/, /g/, /z/, /v/ niet verscherpen: outbound /ˈaʊtbaʊnd/, inbound /ˈɪnbaʊnd/, optimize /ˈɑptɪmaɪz/, downside /ˈdaʊnˌsaɪd/, subscribe /səbˈskɹaɪb/
/æ/ — opener dan de Nederlandse “e”: balancer /ˈbæ.lən.sɚ/, caveat /ˈkæviɑt/
/g/ — een harde plofklank, geen Nederlandse “g”: gateway /ˈɡeɪtˌweɪ/, delegate /ˈdɛlɪɡət/, integrate /ˈɪn.tɪ.ɡɹeɪt/

Zo oefen je met deze video

Luister de hele video één keer zonder te spreken en noteer de woorden die je niet kent.
Spreek zin voor zin na op normale snelheid en herhaal elke zin tot je ritme gelijk is aan dat van de spreker.
Neem jezelf op en vergelijk met het origineel; let daarbij op woorden als dedicate, deploy, consume.

Wat is de Shadowing-techniek?

Shadowing is een wetenschappelijk onderbouwde taalleermethode die oorspronkelijk is ontwikkeld voor professionele tolkentraining en gepopulariseerd door polyglot Dr. Alexander Arguelles. De methode is eenvoudig maar krachtig: je luistert naar native Engelse audio en herhaalt het onmiddellijk hardop — als een schaduw die de spreker volgt met slechts 1–2 seconden vertraging. In tegenstelling tot passief luisteren of grammaticadrills, dwingt shadowing je hersenen en mondspieren om echte spraakpatronen tegelijkertijd te verwerken en te reproduceren. Onderzoek toont aan dat het de uitspraaknauwkeurigheid, intonatie, ritme, verbonden spraak, luisterbegrip en spreekvaardigheid aanzienlijk verbetert — waardoor het een van de meest effectieve methoden is voor IELTS Speaking-voorbereiding en echte Engelse communicatie.

Shadowing-techniek: lees de volledige stap-voor-stap-gids →

Shadowing Practice: Azure App Service and Virtual Network Integration Options - Learn English Speaking with Video

Technical deep dive on Managed Instance for App Service | BRK102

Every Networking Concept Explained | Networking 101 (2026)

Top 5 AI Gateway Use Cases | Solo.io

BAS Bootcamp: Day 6 Setting up IO Internally

Master n8n Triggers in 20 Minutes – Automate Anything

Uber Data Engineering Mock Interview - Ride-Sharing Data Warehouse Schema

Understand Microsoft Entra ID | AZ-104 | Episode 2

Cloud Models - CompTIA Network+ N10-009 - 1.3

OWASP Top 10 Explained

The Simplest Way to Connect Claude and Power BI