Cloud Native Computing Foundation CNCF Webinars, 2 Oct 2018

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Webinar: Rapidly Shipping Rust with Kubernetes & Skaffold

Description

Join us for Kubernetes Forums Seoul, Sydney, Bengaluru and Delhi - learn more at kubecon.io

Don't miss KubeCon + CloudNativeCon 2020 events in Amsterdam March 30 - April 2, Shanghai July 28-30 and Boston November 17-20! Learn more at kubecon.io. The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy, and all of the other CNCF-hosted projects

Modern web application development often faces a tradeoff between safety and application speed. Rust is a powerful language which enables developers to get both: blazing fast applications with C-like performance characteristics, while avoiding many of the common pitfalls and security issues associated with C++ memory management. However, despite the capability to overcome these tradeoffs, Rust is a new language. This means successfully integrating it into your tech stack could cost significant engineering time.

Here, Kubernetes and Skaffold can help developers move quickly and safely, with continuous delivery of code to a highly available container orchestration system. This provides resources for automated testing as well as deployment of the final web application making it possible to test their app and get feedback quickly.

In this talk we will show how automating, building, and pushing a service to a Kubernetes cluster allows for quick iteration:enabling the testing of changes remotely as they are being made. We start by developing a REST service from scratch using Rust. The service will be deployed to Kubernetes on DC/OS using Skaffold.

In this Webinar you will learn about:
- How Rust can help you to write fast and safe (web-)applications
- How developers can leverage Skaffold and Kubernetes for faster development iterations
- Shows the hands-on practice on how to deploy Rust and Skaffold on Kubernetes.

Note: We will also present an extended version “Iterative development on Kubernetes with Rust and Skaffold” of this webinar at KubeCon + CloudNativeCon China in Shanghai on November 13–15, 2018.

A

All right looks like the numbers are stabilizing, we'll go ahead and get started, welcome everyone to today's cmcs webinar, rapidly shipping rest from kubernetes and scaffold and Caitlin Barnard marketing manager at CNCs and I'd like to thank you all for joining us today. So just a few quick housekeeping items before we get started during the webinar you're, not able to speak as an attendee, there is a Q&A box at the bottom of your screen.

A

So if you have any questions throughout the webinar, please drop them in there and then we'll get you as many of those as we can at the end and just as a reminder, the session is being recorded and will be sent out afterward, along with the link to the presentation within about 24 hours, if not sooner, and with that I'll hand over to your to introduce today's presenters. Thank.

B

You so much Caitlin for the kind introduction, and also to the MCS for hosting this webinar before we get started. Actually just a brief note on the title, it's Mike's team that we were just trying to pick as many buzzwords and try to fit as many buffers into the title as possible, but there's actually a story behind that and that actually Shane will probably also tell us a little bit more about it. But it was basically that Jane was playing around for us and he needed someone short.

B

She actually ended up using kubernetes and steffel and I and particular finds are very interesting because kubernetes and scaffold are those clouds of native technologies. Where, typically, we don't think it across this, like a cloud typically cloud native tool and so I find it very interesting to see how kubernetes and scaffold can actually also help those non cloud native tools to become more efficient, and this is why I'm pretty excited about this webinar. And can you just please go to the next slice to something.

B

Was excited about the scissors New York I'm, a technical, lead and engineer motion for data five over admitted I'm awesome with our founder Ben Heintzman I'm also have governing board and a unicorn Bela over at matrix, and so I basically have dealt with cloud native system for at least five years of my life and having that fed I'd like and over to get wrong.

C

Hello, I'm Gaston, it's a French name but then I'm actually from Argentina and I. Also look at mesosphere and I'm on SoCo engineer, I've been working on mesos and so I'm. An Apache messes commander been working on the Mentos and related to things in the ecosystem.

C

For about three years now, before that, I was working at able to, as I was working on, ops works, which is kind of like a I like to call it like chef, obviously, the chef on the cloud, so it's in the same space and a few years we have before that I got the chance to work on EDL's at Google. You some work. So that's how I first got into all this stuff.

C

Now I'm really excited to present about this. It's my first webinar so have some patience and yeah. That's all about me! That's now talk about insane who's, the one who prepared the demonstration that we will be watching at the end of the dock.

D

Hi I'm Shane at I'm, a software engineer at Mesa sphere and be working on CA C, D pipelines, mostly working with kubernetes on DCOs in previous lives. I worked on Rioch at bash, show and infrastructure as a service and software as a service at liquid web in the web hosting industry.

C

Okay, so having that all the interactions and we're ready to get started, I would like first, you give like an outline of what this webinar will. Look like at first I will try to give like a introduction on more development techniques and say why we think that all these passwords in the title are relevant and how they be.

C

The story then point out some of the challenges that people starting using these technologies and following these processes might face, then we will introduce some tools that can help with those changes and, finally, we will use them all together and we will present a flow that can be followed by developers and there's a pre-recorded demo that will play by chain, but he is here available. So if you have any questions for him, you can of course make them at any point.

C

So, having said that, and let's start with the first part of this presentation,.

C

Basically, I'm going to speak about some of the new things that are happening or not so new in space. So in this slide you can see the USO falls, the what the first interaction dinner and so quite some people might be used to, or at least used to do, development or work on projects using the waterfall model, in which you speed your project into different stages. And then you first design your project and you develop it, and then you test it, and you only do those things once and after the previous stage has finished. I.

C

Like to say that a more modern approach- but it's like not that new anymore, it's like 10 years old, it should do your development and continuous or iterative way. So basically, you still design development test in your stuff and normally there's no much overlap between that. But you have been in in cycles. So you decide a bit of your project. You develop that you test it. Yet it works fine that you could be newer, that you're more confident and then you can do an hour, iteration and yeah. Depending on your project.

C

You can have like tens of iterations or more so since you're, going to iterate a lot. It's really important that each iteration is cheap and quick. You cannot spend days doing a deployment or trying your stuff otherwise, and you will really slow down so in order to help doing that to be able to ship quickly and you can use some CIC systems like a continuation from Venus development systems and these systems basically help you automate that process of building your software and deploying that and.

C

Since your this is like a CNC, F webinar and we're going to be talking about containers and how they help make it easier to deploy your applications and yeah how you can use the same stuff while you are developing your your application and then ship at their own production and have your CI see those and build all of that, so, basically using a CIC, it will helps you achieve quick iterations that are necessary if you're, following a more agile model- and our thing that has changed is well it's a something that has changed like it.

C

When you're working on an application, you might go through these stages and at first you might start on a small project with what is kind of kind of like called pets. So you have a few servers that you know very well and you have given them good names and you install all your stuff there manually and usage there and you really care about them. If something goes wrong, you kind of like bring it to the veterinary, you try to fix it and heal it. I can cause you downtime, and so that could be fine.

C

If you have few users or if you're just getting started, but once you go to production or if you are going to scale up your service, then you might want to treat those services so having more servers, and you can treat them as this anymore.

C

You might want to give them warning names like a web server once web server queue or something like that and freedom as cuddle, where you have some tooling like to do some configuration management tooling or you could use fogger to create images or something like that, and then it's really easy to bring up a new server, just like the other ones that you have already, and if one of them is sick and or you know, has some problem and goes down, you can just simply replace it easily that works better when you ya have more users or you need quicker iterations, and the next step is kind of like an abstraction.

C

On top of that, you can see all your servers. That's just one cluster. In this feature you see just like one sheep, on top of which you can just deploy your applications. You pack them your application in a container, and then you have some system, you could be coordinators, you could be Nomad or mesos and that will take those containers and deploy them somewhere in your cluster.

C

One thing that should be you should have in mind is that even though you can think of your classes, just one thing as a sheep or something like that, there are still servers underneath that, so someone will still have to take care of all the all the servers. But that's something it's a topic for another talk.

C

Okay, so if you use an CIC DTOs and then use something like cornets and that stuff and then then you can iterate much faster and but there are still some challenges that you will face some things that are still not super easy and if you're going to use containers, you have to build them each time that you can meet new code or that you write new code, you have to take care of pushing them to either a remote docker registry or locally somewhere, and then you have to do a deployment.

C

You might want to do a quick deployment to in your development environment and that could cause an downtime. But you are not worried about that because you're just testing it yourself, but once you go to production, you will want to and like I, don't know, canneries or Bluegreen deployments stage deployments and stuff like that. So you need some tooling or automation or some process to handle that, and also a common pitfall of moving applications to containers is that you just basically wrap everything.

C

You know all your application and all the dependencies and everything in a single giant container and based on a large distribution like wound, or something that and and then you might end up with a Cheyenne. So you might move from having a giant binary to having a giant container. That's like a three or four gigs large, and that has some problems like healing that takes some time and you need that powerful machine word not super powerful, but you need a decent machine and each time you build something.

C

Well, let's get slow, because you have to fetch all the layers and then run all the steps. And you know you can catch some things. It had some latency to your development process and that latency is also present. When you're deploying your service, you will push you to kubernetes or something like that, and then each pupil at each node will have to pull your image. That can be very large, and so it will take a time for your service to be life and ready, mm-hm and another challenge.

C

This is not so much inner end using containers, orchestration tools or something that it's more something that luckily developers have been focusing more on lately is safety and basically making sure that, if you're using on currency and first you're using those things in a safe way and memory, safety is also pretty important.

C

At least I'm an I spend most of my time working on messes on C++, and it's really easy to shoot yourself on your foot and half I know your application crash or some security problems, but we will introduce in the next slide, actually rust, which can help you with that.

C

So this is how Russ so from so did actually pretty slide. I was introducing some techniques and stuff like that. Now we're going to talk about some tools that you can use to actually create a flow for your developers to use and to be able to iterate quickly and if things rapidly to production and the first tool is rust, it's a programming, language weighted mochila and it was designed to be safe, concurrent and practical.

C

So that means that safety is built in into the language. It's not an afterthought the compiler and then, if yeah, the compiler will check many things for you, while you're writing code. So it's really hard to write, unsafe coding in rust. You basically have to rub your unsafe code in and safe cube, or something like forgot to say is that this is a low-level and Static language. So you get a binary when you compile your your application and that performs really well there.

C

You need a Google machine or interpreter or anything like that to run it, and the performance is similar to idea Matic C++ if you're, right, writing, eggomatic and rust and.

C

Yeah, it has strong emphasis on safety and control the memory layout, so you know that no one know our thread is going to overwrite. What you're writing you? You know who owns each structure that you create and last but not least, there's a great and I'm very friendly community behind it. That's pretty refreshing if you're coming for another language, so yeah I encourage you to play with it. If you can and oh talking about performance again like it, you could compare this to go.

C

One of the things is that rust doesn't have a garbage collection, so in some cases it might perform better.

C

They are, of course, some challenges with rust. No, nothing is perfect, there's no silver bullet and one of the things he said. It's a really deeply young language, it's a few years old, but it's still, you know compared to C++ or C or Java.

C

As I said it brings some extra safety and the compiler is really strict and doesn't let you mess up or tries really hard to prevent you from messing up. But for you to do that, you have to pay some upfront price. Okay, you have to understand some new concepts like what a burrow is, and you have to really get familiar with the language before playing with that and because it's a new or well pretty young language.

C

Many engineers are still pretty new to it and integrating it into a stack Mattamy, not smooth, because they don't have experience with that, and then they also will start experimenting that and getting errors and stuff like that.

C

So they will need to iterate very quickly to be able to get up to speed and one of the errors that I mentioned or things that could happen could be, for instance, trying to write a HelloWorld program and then getting an error from the compiler saying: hey print line doesn't exist, did you mean like Rinda and excavation by that's a macro and because in rust, prin tine is actually a macro? It's not a function and yeah I thought I.

C

Finally, rust compiler is really helpful and gives you a lot of really useful messages, but you have to be able to quickly run it and see where your code is fine or not where it satisfies the compiler or not, and then ehh writer than that, and our kind of message that you can get is like trying to modify an immutable variable in rust where words are immutable by default, and that's also not that common in our languages and that can catch you by surprise again.

C

The compiler is really friendly and helpful, and it will let you know about that, but you need to run it often in order to find your opponents. If you have already read a lot of code, then you will happen. You will spend a few hours just fighting your compiler and fixing each error only to find a new one.

C

So yeah that that was pretty much why we chose rust. We thing as York said that it's not one of the most common cloud native language is used, but it's really valuable in Easter these days and it benefits. If you can iterate quickly, then your experience using it gets much nicer and we will move on to coordinate this and again I.

C

Think that's one of the reasons why you're here you want to look like this and about how you can develop your own coordinate is quickly, but I still do all I can I like to get a short introduction on what it does so cornetist lets.

C

You see all your servers that used to be pets or cattle, ask just one cluster and then using different tools or you can, but you can just use queue, cattle or you can use some tools like scaffold and then depot your or your system, you see in containers or on that cluster and Kronus will take care of orchestrating those containers and choosing, where they'll and making sure that if one of them goes away because of some problem, it is replaced and yeah basically makes your life much easier.

C

And then you can use coordinators remotely when you're deploying on production. But you can also use it locally. You can use something like mini cube or a small development cluster and then use that while you're developing your application, one of the nice things that even else used to use in middle infrastructure. So basically you can build a container once and you just deploy that if you need to scale up scale down, you can just get more instances of your container.

C

You don't have to create new machines and then run any chef recipes or any else will play books or anything like that. You can just add any like scalable containers, your service and in order to let you scale things up and down and do deployments stuff like that. It provides some useful, abstractions and tools, some of which we will use in the demo but still like.

C

If you are deploying an application here using coordinators, you have to take care of building your application of pushing it somewhere of doing a deployment, and you could automate that with it guys. It's a lot of manual work, otherwise and likely. There are some tools that help you automate all that process and one of them is shuffled by Google.

C

It's basically CLI tool that facilitates or make it easy to do, continuous development and using coordinates what it does is it handles the workflow from building your application, so you still have to code it.

C

Of course we are not replaced yet as engineers and but once you have saved or committed your changes, then scaffold can take over and build a new image for you put it somewhere, which one registry could either be locally or remotely and deploy it for you and again, you can use it to iterate on your applications while working locally, so you can just as you'll see in the demo.

C

You can change the line in a line, a line of code in your source and just save it, and then scaffold will be watching your code and notice that you changed something and you will do all the run. All the workfolk like build pitch and deploy push a little boy, and you will see your changes almost instantly or you can use the same tool to deploy own production basically and put your changes to a remote cluster.

C

So as I was saying and scaffold basically takes care of, these three steps will push and deploy it's very configurable. So if you are doing some local development, you can make it and use your local local demon so to run docker build when you are building your software and the push stage would be skipped because you already use your local local demon. So there's no, where to push it there and then you will deploy it using cube cuddle to your mini cube cluster that you're running locally.

C

So that's basically how you can get started business scaffold.

C

Once you are more confident or when you want to keep your stuff into production, then you can change your scaffold configuration and make it use something like Google container build builder for for a good stage.

C

Then, once you have the output from Google container builder, you can push it to Google container registry or it could be just a remote local registry on eCos or any other registry, and then from there you can deploy it on gke or on any any remote cornutus Glasser and it no--don't supports you castle, but it also supports other tools like ham, so it's pretty flexible and basically there are two different modes in which you can run this.

C

We will talk about that in the next slide, but the way in which you can get started using scaffold is to first create your Q&A. This manifest will need that anyway, if your recent corn, it is so that's not an extra overhead that you like it's, not something you have to do for scuffle, but it's very important because it needs to know how to deploy your application. Once you have that written, then you have to create a new file.

C

It's this KML file and there you tell scaffold how to perform each of those meals, push and employment stages, and there you did which tools it should use whether it should locally or remotely and that stuff. And then, when, while you are developing locally on your machine, you can run scaffold, eV and and what scuffle will do it? It will run in the background, actually will be in the only program in your console console, but you can forget about it for awhile and then you can work on your application and each time you save something.

C

Each time you change the file. Scuffle will notice that, because it will be watching a repository for the directory and then it will run that pipeline. That will push deploy pipeline and it will let you what, if there, if there are any errors while building applications, you will notice right away and otherwise we'll perform all of those steps successfully, and then you will be a which is your changes live in seconds.

C

The last step that you want to do in the project is in the red scaffold into your CI CD tuning into Jenkins, or something like that, and so scaffold also has this run mode or run command that will run a whole pipe and only once so, scuffle will not watch for any changes or anything like that. It will try those steps exit if it's successful, but also if there are any any problems.

C

If there's an error, then it will print that lock that and exit as well and with a failure, and so what it lets you do is you can set up Jenkins to watch for any new commits or anything like that on your repository and then when you have something, so you can see b23r scaffold and just do a single run, and that way you can use the same tool that you use to develop locally and to deploy things to production.

C

One of the advantages of scaffold and cornetist and all the stack that were describing said it makes it really easy for someone to set up a development environment, at least in the past and in some projects. I worked at I had to spend like a day or a few days, setting up all the dependencies and all the environment, and if you use scaffold on cue and medicine dollar, then it's really quickly.

C

Just pull on your repository, maybe install mini cube, that's pretty easy and automated and and docker, and then you can just run scaffold and change your project and then you can see your changes deployed live in a matter of minutes or something like that. So this is basically a chief like a flow of what our demo looks like and what the workflow for Glover is on the left. You have a hobby developer. Who is writing a lot of rest? Maybe getting started with rust and and then on local mode.

C

He will save a file and scuffle we'll be watching those files, and then that update will trigger and a scuffle crime, and what scuffle we do is it will build the docker image. It will skip the Bush stage slowly and then it will deploy its Oconee. This cluster, where you can see your changes live. If your end is on CI, then you're happy developer. Instead of saving a file will first save it, try it locally and push it to CI and then Jenkins will be triggered and I will Co scaffold and the rest is same.

C

It scuffle with Global's. In this case it will really push something and.

C

Yeah that was basically my introduction to this, and now we will move to this demo and we didn't want to leave it up to the live demo. Goats. So we have pre-recorded, but Jane is here and I can post it at any any moment this one so yeah I must do that. Now you will see some flicker on the screen. I have to switch what I'm sharing. So this would read with me for a minute and when I do that, you can see github and link on the bottom.

C

So this is all in a public github repository and feel free to visit that and play with all this stuff.

B

B

C

Okay, so I would start Aveda now. Let me know if the audio doesn't work or something that I tried this before and should work with you. No, no. Please.

D

Then we'll start the sake. This is a demonstration of deploying Russ to kubernetes on DCOs, with scaffold in this demonstration. We're gonna build a small rust web app with rocket dot, RS and diesel, which we're gonna have automatically deployed and redeployed on changes for us with scaffold.

D

There are several steps in this demo. First we'll set up kubernetes ingress, then we'll start the base rust application. Then we'll do our first deployment and redeployment with scaffold, then we'll add the database and the REST API, which make up the app and watch all of our changes shipped out to kubernetes automatically by scaffold. In the background. After that, we'll do cleanup in conclusions.

D

For this demonstration you need a dcs 1.11 + cluster, with kubernetes on DCO s130 110, a de greater docker coupe CTL + scaffold.

D

For the purposes of this demonstration, we're going to gloss over installation of DCOs or kubernetes um DCs as those are covered in our documentation, which is linked here. If you're not familiar with these 2's, it has the idea of public versus private agents and kubernetes on DCOs deploys its nodes either to a public or private agent.

D

So for this demo I used a small communities cluster with three public nodes and three private nodes. If you want to run the demo, make sure you at least have one public and one private kubernetes node in your kubernetes on DCOs cluster next we'll be setting up kubernetes ingress for this demo, we'll be using engine X for ingress.

D

All we need to do is run this line which will grow and grab our manifests for engine X ingress. The important one to focus on is the daemon set for the engine. X ingress controller at the bottom here there's a specifically kubernetes on DCOs node selector entry. We are only going to be deploying to our public.

D

So now we can see that I have the engine ex controller running on all of my three public notes.

D

From here, you're gonna want to make sure you export the variable public node IP I've already done that in the background. Basically, that needs to be exported to the public IP address of one of your DCOs public agents, how you get that as dependent on your environment,.

D

Next, we're going to set up our base Russ app and do a small build to test it. First, we'll need to cargo dot. Tamil file, which is the configuration for our dependencies, will be using diesel, which is an ORM for rust, so that we can talk to Postgres later rocket. Is our web framework and there's some other things in here for handling JSON and so forth. We're gonna drop that file in our demo directory as cargo tunnel.

D

Next, we're going to create our main dot RS file.

D

All this is doing for now is printing out rocket web server, we'll be making changes to this and having a scaffold, deploy those changes automatically in the background, but for now that's just for our initial tests.

D

Next, we'll add our docker file there's a couple of important notes that are down here about the docker file. For one we pull a very specific version of Russ Knightly Knightly is required for rocket dot RS, and this is the last version I tested with this demo. We're also going to be doing some tricks in here. This is noted here. Cargo. The build tool for rust does not have a dependencies only build option so to help with backer caching.

D

We do some tricky stuff here and then ultimately, we're using a multi-stage docker build so that we're altima emitting a image based on Alpine latest so that it comes out to less than 20 Meg's, very small image so we'll create that we'll also create a docker ignore file.

D

I've done this in the background, but from here, if you're walking through the demo, you'll want to make sure that you lock and log in to docker haba, okay moving forward we're gonna, do our first test. Do a docker build in the background. I already did the bill just because the first one takes a while and then run the app locally in docker for testing. So it should be up and running as rust web demo, and then you should be able to reach it here.

D

D

Now we can move on to starting to deploy continuously with scaffold, we're eventually going to be adding a postgres container here and we're also going to be making changes to our app and scaffold. Dev mode will run in the background and push those changes up for us. It's going to need manifests for that. So the first thing we're going to do is we're going to create a scaffold deployment da mold, which is gonna, include the ingress service and deployment for our rust web demo.

D

At this point, you'll want to export the variable demo domain. I've done that in the background, you can pretty much choose anything. We're gonna use the host header with curl to actually make the tests against the the cluster, make sure that you have scaffold installed, that you're running the latest release, and then this is our scaffold configuration itself which is just going to say these are where our manifests are and the scaffold appointment file we just made.

D

We'll need to edit both of these files.

D

And make sure you change your user name to your actual user name on docker hub, so mine's Shana do the same on the manifests and also replace demo domain with the demo domain that you're gonna use, so I'm gonna use a non-existent subdomain.

D

Next, we'll do our first deployment was scaffold. We're gonna, run scaffold dev to continuously deploy changes out to our kubernetes cluster. You can add the VD bug flag if you'd like for additional verbose output to either just let you know more, what scaffold is doing or if you're having problems that can be helpful, so we're gonna go ahead and run that go.

A

D

And run that over here and it's gonna go ahead and deploy our rust web demo.

D

And then momentarily there we go should be able to curl it and see Rocket webserver.

D

Next, we'll do our first demonstration of scaffold automatically redeploying, so this is the same main RS file more or less, except for this line, we've changed it from saying rocket web server to saying scaffold updated me. We're gonna, just override the whole file.

D

Scaffold will do its business in the background.

D

And we should just be able to start curling and eventually get the updated version of yeah there. It is. We have three replicas takes enough for them all. Okay,.

D

Next, we're gonna deploy a Postgres database. This is not an H, a one. That's just a demo post past database. We're gonna have scaffold, deploy it for us and manage it. For us. We'll add this: these manifest the service and the deployment for post grafts to our scaffold deployment.

D

The scaffold will go ahead in the background and set up the Postgres server for.

D

And next we'll set up diesel, so diesel is an ORM for us which we're gonna use to talk to post grass. One of the things you want to do in the background is set up this port forward, we're going to use the local diesel CLI to deploy our migrations to the Postgres server. If you want to take this demo further, something like kubernetes init containers would be a better way to do that, but just for demonstration purposes we're just gonna.

D

Do it this way, because it's simple so I have that running in the background, so I should be able to go ahead and talk to my database there. It is.

D

Next, we're going to install the diesel CLI, which is used for managing migrations. I already have it installed, but that's all you need to run is just cargo install diesel, CLI and then say we want the Postgres feature. Then we're gonna create a DMV file which is going to contain the connection. Information for postcards will run diesel setup which will create our migrations directory, get that template it out for us and then we're going to override the initial up-and-down migration. So we're gonna create a table called employees.

D

Employees are gonna, have a first name, a last name, an age and the title.

D

And then our down is our rollback, so we're gonna drop the employees table in the sequence.

D

Once that's done, we can do our initial run of the migration.

D

And then we'll do a redo as well, which will do the rollback and then rerun the initial migration just to make sure that everything is working.

D

Everything worked properly should be able to run this look at the tables and see if the employees table is there.

D

From here we're gonna, add diesel code to manage Postgres connections and talking to the database for us, so we'll add the Postgres that RS file. This is just gonna, have a function that produces a PG connection for us.

D

We're going to add models, dot RS, which will include our employee model.

D

And then we're going to print the schema using the diesel CLI.

D

There we go again, you may want to do this a different way if you take this demo further, but for now we're just gonna manually add a default employee.

D

Update the sequence, and then we should be able to see that we've added our default employee some person, a software engineer, age, 25,.

D

The rust web demo to get the database URL the connection information for Postgres, we're gonna, add a secret, so that is the rust web demo database URL secret. That's just getting tagged on to the end of our scaffold deployment. Then scaffold we'll go ahead and create that secret for us.

D

Now we need to add an environment entry so that the app has access to the database URL so we'll just edit. The.

D

Scaffold appoint itself and tag that on here.

D

Okay, scaffold we'll send that up for us and reconfigured went up there, but it reconfigured the app for us.

D

Now we're gonna do we're gonna update our main dot RS and instead of just printing out scaffold, updated me or rocket web server, we're going to have it actually show us the default employee and pull it from the database. So that's what this new main that RS does. Ultimately, if you get you'll get the default employee and then the results using diesel from the database and.

D

Then, once everything's working should start getting the updated version here, surly.

D

D

There's and then all three of our replicas are done. Okay,.

D

Okay, next we're gonna set up the rest API so that we can get put post and delete the employee data. One thing you can do here and you can actually do above pretty much any time if you want sometimes I get in the habit of just control-z stopping scaffold in the background. I want to running scaffold dev just because, if I'm changing a lot of files, I may not want it to constantly ship out, so we're gonna override our models.

D

That are s, we're going to expand upon the employee model, we're going to derive a bunch of different traits for it, so serialize and deserialize for jason, Coralville and insertable for diesel, and then we're going to create an Employee List which is serializable so that we can just produce a list of all the employees in the system.

D

We're gonna add some just very basic: we have errors. If you take this demo. Further you'll obviously want to expand upon the errors, then we'll create a form for employees so that we can accept employee models via JSON.

D

And then, finally, the meat and potatoes of the demo, we're gonna, add all of our get put post and delete methods. So this API dot RS will have a 404 catcher and then get employees which will produce employee lists, get employees with a specific employee ID, which will give you the information about that employee. Put employee for adding a new employee to the system post to update existing employees and delete to remove an employee from the system.

D

So next we're gonna override our main dot RS and bring it all together. We're going to deploy rocket with the routes that we build above and start testing yeah.

D

Okay, we can bring scaffold back and let it do its job.

D

Now everything should be working, we should be able to get our default employee via the API.

D

There's some person aged 25 software engineer get a list of our employees just that one. For now we'll add a new employee. This will be new person aged 27, DevOps engineer.

D

If we go and list you'll see that they're now added.

D

We'll update the new employee, so let's say there was a clerical error and we are going to update their age from 27 to 29. Their ID is 2.

D

And then we'll take a look at the employees again we can see both of them are in there. The changes, they're, no longer age, 27, their age, 29 and then ultimately delete and delete our new one and the original one. And then, if we lists nothing in there, that's pretty much it for this demo at the end. Here, if you want to clean up, you can press ctrl C once and scaffold will take care of cleaning everything up that it deployed for you and then to clean up the rest of the resources.

D

We added um just do a delete on the ingress engine xtml that we originally created from.

D

And that's pretty much it so there's some notes here in the conclusions about where you can take this demo. Next, if you'd, like things like using an H, a Postgres setup using diesel connection, pooling and rocket, manage state all links here and links to scaffold and Mesa sphere, kubernetes and eCos and the DCOs documentation. Thank you.

C

Okay, so that was basically at the end. Jen mentioned some things that you could do. If you want to take this demo further, and just repeat that we really don't recommend running Posterous like there, you can run anything on production. You might want to make sure that you run an H, a high-level poster setup. We stand for volumes so that, if you do another deployment, you lose your data, and so you can buy that do backups and stuff like that and and yeah.

C

There are some features from the rust libraries that he didn't use for this demo and you might want to experiment with them with connection pools and using manage stating rocket stuff.

C

Yes, we also want to let you know that we will be presenting on this topic at Yukon China, so we will have add more content and make it an interesting, and basically we will do it. Reddit development on this demo as well and in this presentation and yeah we'll be presenting in China on when we're before games, I, think something that so I look forward to seeing you there meeting you there and now it's time for questions. If you have any questions, please.

B

Thank you guys so much for presenting, especially the demo. I was really excited and if I hadn't actually run through it this morning, I would probably love to see the next thing. I would do right out there in terms of question. We don't have too much time left. So I would actually just asked one question, and that would be what would be kind of alternate suspect. You can come up with this smell like one particular choice of technologies, but what other options.

C

So I think you can use scaffold for any language.

C

Someone was saying that he was interested in using it for nodejs and go basically at the build step you can make. It run, use any docker image. So if you change the docker file, you could be building a go application and about application and even support space as well. So you can switch the language to whatever you prefer and, and everything else will work. The same and I would still use scaffold.

C

Then, because it's very nice, it watches your changes and it will do all this for you, but you can switch also your class or again instead of using these new OSU Canada people that GAE or.

C

D

I was gonna, say something similar so like the demo, I wrote it originally because I had a kubernetes and eCos cluster available. This demo works just as easily on mini coop and in fact, scaffold will automatically do port forwards for you, so you don't even have to use any grass or anything like that, but pretty much you know take out any peace you want scaffold doesn't even have to be part of an actual like a full-on CDI CD pipeline. You can just use it for development on mini coop or something like that.

D

That's a good use of it as well, so so.

B

You would even give your local development. The same style is just a mini, cube cluster and sells like a lot of structure somewhere else as they're deploying yes,.

D

B

All right, thank you so much again for presenting feel free to follow up with questions afterwards and also hope to see some of you at keep on China. With that out hand back to Caitlyn he's going to do the housekeeping and finalizing.

A

Awesome thanks so much for the great presentation today and thank you. Everyone who joined us for today's webinar just a reminder. The webinar, recording and slides will be online later today and you'll receive an Inc link directly to your email. So, looking forward to seeing you all at a future, CNC F webinar have a great day. Everyone.