[00:00.000 --> 00:05.000]  Welcome, Bobor.
[00:05.000 --> 00:09.000]  Hi, everyone. Thank you.
[00:09.000 --> 00:14.000]  Thank you. Hi, everyone.
[00:14.000 --> 00:16.000]  I am super excited to be here.
[00:16.000 --> 00:19.000]  Welcome to one of my sessions about
[00:19.000 --> 00:23.000]  observing your APIs with API gateway plugins.
[00:23.000 --> 00:25.000]  Let me introduce myself first.
[00:25.000 --> 00:27.000]  My name is Bobor.
[00:27.000 --> 00:30.000]  I am a developer advocate for Apache API 6.
[00:30.000 --> 00:33.000]  Sometimes it's so difficult to pronounce my name.
[00:33.000 --> 00:35.000]  People say it's from different countries.
[00:35.000 --> 00:38.000]  Is it like a Bebo or Bobor or like a Bebo?
[00:38.000 --> 00:43.000]  And then I say, okay, you can translate my first name as a tiger.
[00:43.000 --> 00:45.000]  It means tiger, Bobor.
[00:45.000 --> 00:49.000]  My last name, Murzokov, English version will be Livermore.
[00:49.000 --> 00:51.000]  In this case, Tiger Livermore.
[00:51.000 --> 00:53.000]  You can call me like Tiger Livermore.
[00:53.000 --> 00:55.000]  It's up to you.
[00:55.000 --> 00:58.000]  And you can also reach out to me on these social channels
[00:58.000 --> 01:01.000]  if you have any questions regarding sessions.
[01:01.000 --> 01:04.000]  So with that, we can get started.
[01:04.000 --> 01:07.000]  First thing first, what I want to do now is take a selfie
[01:07.000 --> 01:10.000]  because I have my TikTok account.
[01:10.000 --> 01:13.000]  I just start to run my blog recently.
[01:13.000 --> 01:18.000]  Just a moment. Maybe this is better.
[01:18.000 --> 01:21.000]  That was good. Thank you.
[01:21.000 --> 01:25.000]  I will just put some hashtags on that.
[01:25.000 --> 01:27.000]  So thank you Fabiano. Now I can leave.
[01:27.000 --> 01:30.000]  I did my job. I can go home.
[01:30.000 --> 01:33.000]  I did my Instagram picture, right?
[01:33.000 --> 01:37.000]  But I have today a very interesting agenda for you.
[01:37.000 --> 01:41.000]  We'll talk about what is APIs and API observability
[01:41.000 --> 01:47.000]  and how we can use API gateway for observing your APIs
[01:47.000 --> 01:49.000]  as a central point for observation.
[01:49.000 --> 01:53.000]  And then we will break down all API observability three pillars.
[01:53.000 --> 01:58.000]  We know that we have logging, tracing, and metrics, right?
[01:58.000 --> 02:01.000]  And we will learn how to enable these three pillars
[02:01.000 --> 02:04.000]  by using Apache API 6 plugins.
[02:04.000 --> 02:07.000]  And I have then a small demo for you.
[02:07.000 --> 02:09.000]  I hope you will like it.
[02:09.000 --> 02:11.000]  And that's good.
[02:11.000 --> 02:12.000]  APIs, right?
[02:12.000 --> 02:14.000]  API is just the three letters
[02:14.000 --> 02:16.000]  acronym for application program interface.
[02:16.000 --> 02:19.000]  By now, we are all familiar with this term, right?
[02:19.000 --> 02:24.000]  Because we are living in increasingly API-centric world.
[02:24.000 --> 02:27.000]  Even chat GPT uses API.
[02:27.000 --> 02:31.000]  Because under the hood, it calls some OpenAI API lists
[02:31.000 --> 02:34.000]  to collect some language models.
[02:34.000 --> 02:36.000]  Now my question to you.
[02:36.000 --> 02:40.000]  Who doesn't know what is API?
[02:40.000 --> 02:42.000]  Everybody knows. You don't know.
[02:42.000 --> 02:45.000]  You are lucky I have a brief for you.
[02:45.000 --> 02:50.000]  You will get this T-shirt because you don't know what's API 6.
[02:50.000 --> 02:53.000]  What is your size?
[02:53.000 --> 02:56.000]  I will handle it for you here.
[02:56.000 --> 02:58.000]  Actually, I have enough T-shirts.
[02:58.000 --> 03:01.000]  If you are first the three good,
[03:01.000 --> 03:04.000]  if I get the first three good answer, I will give it to you.
[03:04.000 --> 03:08.000]  And we have also some medals, stickers if you would like to get,
[03:08.000 --> 03:09.000]  please feel free.
[03:09.000 --> 03:13.000]  It's like free stuff from community, I would say.
[03:13.000 --> 03:17.000]  And we know that success of your services, right,
[03:17.000 --> 03:22.000]  depends on the performance, availability, and integrity of your APIs.
[03:22.000 --> 03:24.000]  Here, another question, Rice.
[03:24.000 --> 03:28.000]  How to achieve these three indicators of success?
[03:28.000 --> 03:32.000]  How to achieve these three indicators of success for your APIs?
[03:32.000 --> 03:35.000]  Let's say we have API should be all the time available, right?
[03:35.000 --> 03:37.000]  It should be integratable.
[03:37.000 --> 03:40.000]  It should be performance with a high performance.
[03:40.000 --> 03:43.000]  Do you have any idea?
[03:43.000 --> 03:47.000]  Yes, you got one T-shirt.
[03:47.000 --> 03:49.000]  Yes, monitoring, exactly.
[03:49.000 --> 03:52.000]  And what do you say, distributing?
[03:52.000 --> 03:55.000]  Yeah, distribute system, monitoring?
[03:55.000 --> 04:00.000]  Just placing it as a service in a distributed process.
[04:00.000 --> 04:01.000]  Yeah, this is also a rice solution.
[04:01.000 --> 04:04.000]  But you will get only the medal.
[04:04.000 --> 04:05.000]  Sorry.
[04:05.000 --> 04:09.000]  But one solution can be using by API gateways.
[04:09.000 --> 04:13.000]  Because API gateways, actually, nowadays, what we do,
[04:13.000 --> 04:17.000]  we build multiple microservices and maybe several less APIs
[04:17.000 --> 04:23.000]  or multiple REST APIs for our unique, maybe, the service, right?
[04:23.000 --> 04:27.000]  And in this flow, API gateways serves as a central point
[04:27.000 --> 04:32.000]  for routing all your incoming requests to the internet destinations.
[04:32.000 --> 04:35.000]  These destinations can be, as you can see, database,
[04:35.000 --> 04:38.000]  or maybe short-party API services.
[04:38.000 --> 04:42.000]  Or it can be also some serverless APIs like Azure Function
[04:42.000 --> 04:47.000]  or AWS Lambda, maybe any other open source functions, right?
[04:47.000 --> 04:51.000]  And it means it's accessed a single layer between your clients
[04:51.000 --> 04:54.000]  and the backend services, right?
[04:54.000 --> 04:58.000]  That can manage all the traffic coming to your backend services.
[04:58.000 --> 05:00.000]  It's a very straightforward term, right?
[05:00.000 --> 05:01.000]  API gateway term.
[05:01.000 --> 05:06.000]  And then it can also be a right point to learn
[05:06.000 --> 05:11.000]  for your API observabilities because it's uniquely identified
[05:11.000 --> 05:14.000]  to know all the traffic moving from client side
[05:14.000 --> 05:16.000]  to our backend service network.
[05:16.000 --> 05:21.000]  And instead of relying on, let's say, some other technologies,
[05:21.000 --> 05:26.000]  SDKs, APIs, and services to enable this observability
[05:26.000 --> 05:28.000]  and improve this observability,
[05:28.000 --> 05:31.000]  you can easily integrate this job with API gateway.
[05:31.000 --> 05:33.000]  We have a bunch of API gateways.
[05:33.000 --> 05:36.000]  I'm not selling anything, but we are talking about now
[05:36.000 --> 05:41.000]  API 6, what kind of plugins you can use today.
[05:41.000 --> 05:46.000]  And next, for example, you can ask, what is Apache API 6?
[05:46.000 --> 05:50.000]  Maybe you know the world's largest open source software
[05:50.000 --> 05:52.000]  software foundation, Apache software foundation.
[05:52.000 --> 05:55.000]  Maybe some of you are part of it, right?
[05:55.000 --> 05:57.000]  Who is a part of it, ASF?
[05:57.000 --> 06:00.000]  Who is contributing to open source projects of Apache
[06:00.000 --> 06:02.000]  software foundation?
[06:02.000 --> 06:05.000]  You, we have some people.
[06:05.000 --> 06:10.000]  And who knows Apache Kafka, Cassandra, Tomcat?
[06:10.000 --> 06:11.000]  Everybody knows.
[06:11.000 --> 06:15.000]  And Apache API 6 is also one of the top fastest growing
[06:15.000 --> 06:17.000]  project of ESF nowadays.
[06:17.000 --> 06:20.000]  You can, of course, you cannot compare it with Cassandra,
[06:20.000 --> 06:22.000]  Tomcat, but maybe in the future.
[06:22.000 --> 06:25.000]  It was initiated in 2019, but we are still,
[06:25.000 --> 06:27.000]  we have some open source community around the world.
[06:27.000 --> 06:30.000]  I am, for example, visiting Tallinn, Estonia.
[06:30.000 --> 06:33.000]  We have contributors from US, Canada.
[06:33.000 --> 06:36.000]  We have some with China, India, and so on.
[06:36.000 --> 06:37.000]  You can check it out.
[06:37.000 --> 06:41.000]  It's a very nice API gateway solution.
[06:41.000 --> 06:46.000]  And as you can see, API plugins, it's a very hard mechanism
[06:46.000 --> 06:50.000]  in API gateway that can be plugged into your API gateway
[06:50.000 --> 06:51.000]  solution.
[06:51.000 --> 06:54.000]  With that, you can extend further some functionalities.
[06:54.000 --> 06:58.000]  You can enable cross-captain constants like authorization,
[06:58.000 --> 07:02.000]  authentication, security, transformation, rate limiting,
[07:02.000 --> 07:03.000]  and so on.
[07:03.000 --> 07:07.000]  At the same time, you can enable some kind of observation,
[07:07.000 --> 07:08.000]  right?
[07:08.000 --> 07:11.000]  And when you're using API 6, for example, you should face
[07:11.000 --> 07:15.000]  with multiple types of plugins broken down into several
[07:15.000 --> 07:19.000]  categories, or sometimes you want something custom
[07:19.000 --> 07:20.000]  plugin, right?
[07:20.000 --> 07:22.000]  To fit your needs.
[07:22.000 --> 07:26.000]  And with API 6, similar API gateway nowadays, more than
[07:26.000 --> 07:29.000]  API gateway provides some language support.
[07:29.000 --> 07:31.000]  You can choose your favorite language, your familiar
[07:31.000 --> 07:34.000]  reach, and you can create some custom plugins.
[07:34.000 --> 07:38.000]  Maybe you are Java developer, or maybe you are Go developer
[07:38.000 --> 07:41.000]  or Python, you can choose your favorite language and write
[07:41.000 --> 07:42.000]  the plugins.
[07:42.000 --> 07:43.000]  Let me show you.
[07:43.000 --> 07:46.000]  Or you don't want to write a code.
[07:46.000 --> 07:51.000]  There is a dashboard where you can do user-friendly
[07:51.000 --> 07:52.000]  dashboard by using it.
[07:52.000 --> 07:56.000]  You can just drag and drop existing plugins together to
[07:56.000 --> 07:58.000]  build new plugins.
[07:58.000 --> 08:01.000]  You can orchestrate one or multiple plugins in this way.
[08:01.000 --> 08:06.000]  You can specify some conditions and also put some
[08:06.000 --> 08:09.000]  business flow on it and generate new plugins.
[08:09.000 --> 08:13.000]  It's a very useful feature that API 6 dashboard provides.
[08:13.000 --> 08:17.000]  For example, you can put together multiple observability
[08:17.000 --> 08:21.000]  plugins in one flow and then maybe enable full
[08:21.000 --> 08:25.000]  observability backend tools in this diagram.
[08:25.000 --> 08:28.000]  Let me, you can just maybe observe later.
[08:28.000 --> 08:31.000]  Now let's jump into the main topic.
[08:31.000 --> 08:33.000]  What's the API observability?
[08:33.000 --> 08:38.000]  Can anyone tell what's the API observability?
[08:38.000 --> 08:40.000]  In one word.
[08:40.000 --> 08:42.000]  No?
[08:42.000 --> 08:44.000]  No.
[08:44.000 --> 08:48.000]  The other problem, like relation and latency, or
[08:48.000 --> 08:51.000]  just great status codes.
[08:51.000 --> 08:53.000]  Traces, yes, right?
[08:53.000 --> 08:58.000]  And for, I think I gave to you this short, right?
[08:58.000 --> 09:00.000]  For your, yeah.
[09:00.000 --> 09:03.000]  The actual thing is doing is very important.
[09:03.000 --> 09:05.000]  Yes, it's also right, yeah.
[09:05.000 --> 09:09.000]  And it means API observability is all about how you absorb
[09:09.000 --> 09:10.000]  your APIs, right?
[09:10.000 --> 09:14.000]  Instead of relying on predetermined data like metrics,
[09:14.000 --> 09:18.000]  monitoring, and wait for the failure, you can use
[09:18.000 --> 09:22.000]  API observability to enable announce,
[09:22.000 --> 09:26.000]  announce, or announce, in this diagram as you can see.
[09:26.000 --> 09:30.000]  Because compared to traditional API monitoring,
[09:30.000 --> 09:33.000]  we have traditional API monitoring and observability, right?
[09:33.000 --> 09:37.000]  This is monitoring focused on analyzing
[09:37.000 --> 09:39.000]  now and announce, what does it mean?
[09:39.000 --> 09:42.000]  Now and announce means you know what the measure,
[09:42.000 --> 09:45.000]  you know the number of requests, you know number of errors,
[09:45.000 --> 09:47.000]  that you know what the measure, right?
[09:47.000 --> 09:52.000]  But on the other hand, API observability lets you
[09:52.000 --> 09:56.000]  analyze exactly what was the issue,
[09:56.000 --> 10:01.000]  and how this issue occurred by learning three metrics, right?
[10:01.000 --> 10:05.000]  You know, again, like logging and metrics and tracing.
[10:05.000 --> 10:10.000]  So because API observability nowadays is a part of
[10:10.000 --> 10:14.000]  every API development teams, let's say, for example,
[10:14.000 --> 10:17.000]  who can use API observability during the API development?
[10:17.000 --> 10:20.000]  Yes, you can see, for example, product managers,
[10:20.000 --> 10:25.000]  they can use to learn consumption and usage of your APIs.
[10:25.000 --> 10:28.000]  Or maybe security team, they can use to protect
[10:28.000 --> 10:34.000]  or maybe detect some possible API threats from outside, right?
[10:34.000 --> 10:39.000]  So as I said, let's have a look at these three pillars now
[10:39.000 --> 10:44.000]  and learn what API gateway can provide as a solution
[10:44.000 --> 10:48.000]  for these pillars, tracing, logging, and metrics.
[10:48.000 --> 10:52.000]  If I start with logging, it's a very tribal step, right?
[10:52.000 --> 10:55.000]  To start your observability because you can use logs
[10:55.000 --> 11:00.000]  to identify or debug what's happening.
[11:00.000 --> 11:04.000]  If you are a junior developer, you will first start digging into logs
[11:04.000 --> 11:06.000]  because in order to understand the project,
[11:06.000 --> 11:08.000]  because the project has zero documentation,
[11:08.000 --> 11:10.000]  you need to have a look at logs,
[11:10.000 --> 11:13.000]  you need to have a look at some API things.
[11:13.000 --> 11:19.000]  So here, in order to enable this logging,
[11:19.000 --> 11:21.000]  you can use some kind of plugins.
[11:21.000 --> 11:25.000]  For example, API 6 provides HTTP-logger plugin, Kafka-logger,
[11:25.000 --> 11:29.000]  or it depends on which kind of integration you have.
[11:29.000 --> 11:32.000]  For example, you can use Google Cloud Logger.
[11:32.000 --> 11:35.000]  Let's say HTTP-logger is a basic logger.
[11:35.000 --> 11:40.000]  Let's just enable to push all the log data from your API gateway
[11:40.000 --> 11:44.000]  to HTTP-S or HTTP servers.
[11:44.000 --> 11:49.000]  It means you can further drive some useful metrics from that logs.
[11:49.000 --> 11:52.000]  Or let's say, what about metrics?
[11:52.000 --> 11:57.000]  Metrics are actually just the data collected
[11:57.000 --> 11:59.000]  over the time, right?
[11:59.000 --> 12:01.000]  Times gas, but the metrics collected.
[12:01.000 --> 12:04.000]  But metrics you can use in the future
[12:04.000 --> 12:07.000]  in the distributed systems like Elasticsearch.
[12:07.000 --> 12:09.000]  You can acquire these metrics using Elasticsearch,
[12:09.000 --> 12:14.000]  or maybe you can show these metrics by using a dashboard,
[12:14.000 --> 12:15.000]  like a Prometheus, right?
[12:15.000 --> 12:18.000]  The Prometheus dashboard provides some metric analysis.
[12:18.000 --> 12:23.000]  And for these, all the external popular platforms,
[12:23.000 --> 12:26.000]  let's say, like Grafana or Prometheus,
[12:26.000 --> 12:31.000]  Apache PR6 has pre-built connectors, I would say.
[12:31.000 --> 12:33.000]  And for example, like Zipkin.
[12:33.000 --> 12:35.000]  For tracing, we have Zipping plugin.
[12:35.000 --> 12:39.000]  You can just enable it with just one click,
[12:39.000 --> 12:47.000]  and it starts to learn some metrics or traces and so on.
[12:47.000 --> 12:50.000]  So now, enough talking, right?
[12:50.000 --> 12:53.000]  I can show something in real.
[12:53.000 --> 12:55.000]  Because we have a bunch of plugins,
[12:55.000 --> 12:58.000]  as I said, for the API observability,
[12:58.000 --> 13:01.000]  this time I decided to choose, for my demo,
[13:01.000 --> 13:03.000]  Prometheus plugin.
[13:03.000 --> 13:05.000]  Because with respect to Fabian,
[13:05.000 --> 13:09.000]  and I really like Prometheus and Grafana integration,
[13:09.000 --> 13:11.000]  now I will show you how you can easily enable
[13:11.000 --> 13:14.000]  this observability very fast.
[13:14.000 --> 13:16.000]  For that, I mean, you can have a look.
[13:16.000 --> 13:19.000]  That is my report, api6.net docker.
[13:19.000 --> 13:23.000]  It shows all the use cases of API GitHub in one report.
[13:23.000 --> 13:27.000]  And it has a branch called the API observability.
[13:27.000 --> 13:29.000]  If you navigate this branch,
[13:29.000 --> 13:33.000]  you will see a real example of some plugins,
[13:33.000 --> 13:37.000]  how to enable it, and you can have a hands-on exercises.
[13:37.000 --> 13:41.000]  Now, let me switch to my VS code.
[13:41.000 --> 13:43.000]  I'm using the VS code today for the session.
[13:43.000 --> 13:48.000]  But I'm talking about this repository.
[13:48.000 --> 13:50.000]  It has a five branch.
[13:50.000 --> 13:52.000]  You can learn some of the branch,
[13:52.000 --> 13:54.000]  like how you can enable health check,
[13:54.000 --> 13:57.000]  start from health check, and API observability.
[13:57.000 --> 13:59.000]  This is the starting point for you.
[13:59.000 --> 14:03.000]  And then it's very fast to spin up this project,
[14:03.000 --> 14:06.000]  because it uses docker-compose.
[14:06.000 --> 14:11.000]  And we are using, for the backend, api6.net,
[14:12.000 --> 14:14.000]  you can use Java project.
[14:14.000 --> 14:16.000]  It doesn't matter. You can use Python.
[14:16.000 --> 14:18.000]  I'm actually a Java developer,
[14:18.000 --> 14:23.000]  but I like to encourage myself to learn new languages.
[14:23.000 --> 14:25.000]  I start to learn C-sharp,
[14:25.000 --> 14:28.000]  and this small project on .net.
[14:28.000 --> 14:32.000]  So if you do some docker-compose app,
[14:32.000 --> 14:34.000]  docker-api6, of course,
[14:34.000 --> 14:36.000]  it will bring some containers, right?
[14:36.000 --> 14:38.000]  All useful containers, as you can see,
[14:38.000 --> 14:40.000]  like Prometheus, Grafana.
[14:40.000 --> 14:43.000]  We have a product API that is with .net.
[14:43.000 --> 14:46.000]  I have a small endpoint that maybe returns
[14:46.000 --> 14:49.000]  some list of products when you call this endpoint.
[14:49.000 --> 14:52.000]  As you can see, it's very simple here.
[14:52.000 --> 14:55.000]  When I do this, it returns...
[14:55.000 --> 14:58.000]  Let me maybe make it bigger here.
[14:58.000 --> 15:00.000]  It returns a MacBook price
[15:00.000 --> 15:02.000]  and some other product price.
[15:02.000 --> 15:03.000]  Simple.
[15:03.000 --> 15:06.000]  And also, I'm using Windows on my machine,
[15:06.000 --> 15:11.000]  some of the necessary containers are up and running
[15:11.000 --> 15:13.000]  in one kind of docker-compose app.
[15:13.000 --> 15:17.000]  Then if you open the project code
[15:17.000 --> 15:21.000]  on your favorite IDEA or IDEA tool, right?
[15:21.000 --> 15:24.000]  In my case, let's say VS Code,
[15:24.000 --> 15:27.000]  and you can see on the Commons folder
[15:27.000 --> 15:29.000]  some common line examples
[15:29.000 --> 15:31.000]  that I'm going to show today.
[15:31.000 --> 15:35.000]  First thing, how you can enable this Grafana
[15:35.000 --> 15:37.000]  or Prometheus, Zipkin, and so on.
[15:37.000 --> 15:41.000]  With API Gateway, you need to create your first upstream.
[15:41.000 --> 15:43.000]  If you navigate this here,
[15:43.000 --> 15:45.000]  you can see some kind of command.
[15:45.000 --> 15:48.000]  Of course, you can use a dashboard to create this upstream.
[15:48.000 --> 15:49.000]  It's up to you.
[15:49.000 --> 15:53.000]  If you're like a hard worker developer like me,
[15:53.000 --> 15:55.000]  you can just use the kind of commands.
[15:55.000 --> 15:57.000]  What we are doing here, as you can see,
[15:57.000 --> 15:59.000]  I had a product API, right?
[15:59.000 --> 16:02.000]  And then I am creating upstream.
[16:02.000 --> 16:07.000]  And upstream means just a set of backend API endpoints.
[16:07.000 --> 16:10.000]  And I have one single node, one node.
[16:10.000 --> 16:11.000]  You can have multiple nodes.
[16:11.000 --> 16:15.000]  You can have multiple instance of the same product API.
[16:15.000 --> 16:21.000]  For the simple case, I'm just creating this one upstream
[16:21.000 --> 16:22.000]  and one node.
[16:22.000 --> 16:24.000]  Let's jump into maybe terminal.
[16:24.000 --> 16:29.000]  I can open some new terminal here to run this code.
[16:29.000 --> 16:33.000]  So I'm using VSL,
[16:33.000 --> 16:37.000]  because on Windows it's a little bit difficult to run a Linux code.
[16:37.000 --> 16:41.000]  So let me open maybe one more terminal on the Ubuntu.
[16:41.000 --> 16:43.000]  There we go.
[16:43.000 --> 16:44.000]  I hope it's visible.
[16:44.000 --> 16:49.000]  And if I just click and press Enter,
[16:49.000 --> 16:51.000]  now I set API 6.
[16:51.000 --> 16:53.000]  Please create the upstream service.
[16:53.000 --> 16:58.000]  Register my ESP.NET Web API as a backend service.
[16:58.000 --> 17:01.000]  And it should have this kind of configuration.
[17:01.000 --> 17:04.000]  Then next step, next easy step,
[17:04.000 --> 17:07.000]  what I do is I need to create,
[17:07.000 --> 17:10.000]  let's say for prometers, I need to create a row.
[17:10.000 --> 17:14.000]  Because API Gateway has three very basics,
[17:14.000 --> 17:16.000]  like you need to create the upstream,
[17:16.000 --> 17:17.000]  you need to create a row,
[17:17.000 --> 17:18.000]  and enable some plugins.
[17:18.000 --> 17:20.000]  In my case, prometers plugins, right?
[17:20.000 --> 17:24.000]  As you can see, I have some plugins configuration on the top.
[17:24.000 --> 17:25.000]  Only single prometers.
[17:25.000 --> 17:28.000]  And I'm giving reference to the upstream
[17:28.000 --> 17:30.000]  that in the previous step we created.
[17:30.000 --> 17:31.000]  And that's all.
[17:31.000 --> 17:34.000]  Like I'm saying also, URI,
[17:34.000 --> 17:40.000]  for the row to find out which URI this plugin should absorb.
[17:40.000 --> 17:43.000]  I'm saying fresh API slash products.
[17:43.000 --> 17:44.000]  Here we go.
[17:44.000 --> 17:49.000]  And then if I get this command and press and put it to terminal,
[17:49.000 --> 17:52.000]  now I will enable this plugin very fast.
[17:52.000 --> 17:55.000]  An API six admin is saying,
[17:55.000 --> 17:58.000]  okay, now you have a row, you have upstream,
[17:58.000 --> 17:59.000]  now it's time to test.
[17:59.000 --> 18:03.000]  Prometers plugin, I enabled like this couple of steps, right?
[18:03.000 --> 18:06.000]  You see the five seconds or six seconds, it's all.
[18:06.000 --> 18:08.000]  But compared to Java projects,
[18:08.000 --> 18:10.000]  how you enable prometers also,
[18:10.000 --> 18:12.000]  maybe a little bit compared to the same steps.
[18:12.000 --> 18:16.000]  But this API Gateway, you are just extracting
[18:16.000 --> 18:19.000]  a little bit of huge work to separate service.
[18:19.000 --> 18:21.000]  And it's highly scalable.
[18:21.000 --> 18:24.000]  That's one of the advantages of using API Gateway, right?
[18:24.000 --> 18:26.000]  For your observability.
[18:26.000 --> 18:30.000]  And next, for example, I can generate some metrics
[18:30.000 --> 18:34.000]  by calling my maybe API into point several times,
[18:34.000 --> 18:36.000]  maybe like this.
[18:36.000 --> 18:40.000]  As you can see, it's responding me with the product list
[18:40.000 --> 18:41.000]  in the response.
[18:41.000 --> 18:43.000]  Maybe I can do it one more time.
[18:43.000 --> 18:46.000]  So we have some data on it.
[18:46.000 --> 18:51.000]  And now, if I navigate to...
[18:51.000 --> 18:54.000]  Maybe if I can request all the prometers metrics
[18:54.000 --> 18:57.000]  to see some result, right, in this output,
[18:57.000 --> 19:01.000]  I can run another command, maybe like this.
[19:01.000 --> 19:02.000]  Here we go.
[19:02.000 --> 19:04.000]  As you can see, metrics are enabled.
[19:04.000 --> 19:06.000]  I can see some metrics, HTTP metrics,
[19:06.000 --> 19:09.000]  plus some API six metrics, as you can see.
[19:09.000 --> 19:12.000]  If you have API six HTTP status,
[19:12.000 --> 19:16.000]  it was 200 returned and was fine.
[19:16.000 --> 19:21.000]  And, of course, it looks like a little bit ugly, right?
[19:21.000 --> 19:25.000]  Now, you can see even maybe these metrics on the dashboard.
[19:25.000 --> 19:29.000]  Just navigate to localhosts-targets,
[19:29.000 --> 19:31.000]  because we are running on localhost.
[19:31.000 --> 19:33.000]  When you deploy maybe to the cloud,
[19:33.000 --> 19:35.000]  you will have a domain and so on.
[19:35.000 --> 19:39.000]  But as you can see on the prometers, if I refresh it,
[19:39.000 --> 19:43.000]  you will have soon now some metrics on it.
[19:43.000 --> 19:48.000]  Of course, you can specify some parameters for your metrics.
[19:48.000 --> 19:49.000]  And let me...
[19:49.000 --> 19:51.000]  We need to maybe a little bit wait
[19:51.000 --> 19:56.000]  to be enabled these metrics on the UI.
[19:56.000 --> 19:57.000]  Yeah.
[19:57.000 --> 19:59.000]  Let me check the targets.
[19:59.000 --> 20:01.000]  It's unhealthy.
[20:01.000 --> 20:06.000]  Maybe we need to do some local compost down and up.
[20:06.000 --> 20:08.000]  Every time when it comes to demo, something fails.
[20:08.000 --> 20:11.000]  It's unexpected, not expected.
[20:11.000 --> 20:13.000]  Docker compost down.
[20:13.000 --> 20:16.000]  Maybe I will do this trick.
[20:16.000 --> 20:22.000]  And maybe then an up in a moment.
[20:22.000 --> 20:23.000]  Yep.
[20:23.000 --> 20:24.000]  And here we go.
[20:24.000 --> 20:31.000]  If I do up, it will bring and refresh some Docker stuff.
[20:31.000 --> 20:33.000]  This Docker always is kind of a issue.
[20:33.000 --> 20:37.000]  Now, if I do this, targets should be...
[20:37.000 --> 20:38.000]  No, no.
[20:38.000 --> 20:39.000]  It's still yes.
[20:39.000 --> 20:40.000]  It's stopping.
[20:40.000 --> 20:41.000]  Yes.
[20:41.000 --> 20:42.000]  Now we are in an upstate.
[20:42.000 --> 20:45.000]  And then let me also generate some logs once again.
[20:45.000 --> 20:48.000]  Or even it should be visible now.
[20:48.000 --> 20:50.000]  Let me go.
[20:50.000 --> 20:53.000]  API 6.
[20:53.000 --> 20:57.000]  HTTP status, if I have any.
[20:57.000 --> 21:02.000]  Or even you can see any traces, any metrics here.
[21:02.000 --> 21:05.000]  As you can see, I can even see ETCD.
[21:05.000 --> 21:08.000]  ETCD has a storage for the API 6.
[21:08.000 --> 21:12.000]  What kind of data exchanging between API 6 and the storage?
[21:12.000 --> 21:13.000]  And so on.
[21:13.000 --> 21:16.000]  I mean, you can filter out all the metrics you want.
[21:16.000 --> 21:18.000]  And even you can see in the graph.
[21:18.000 --> 21:20.000]  If it's not enough on a promises dashboard,
[21:20.000 --> 21:22.000]  you can connect with Grafana, right?
[21:22.000 --> 21:24.000]  Very easy.
[21:24.000 --> 21:27.000]  We have an API 6 Grafana dashboard.
[21:27.000 --> 21:32.000]  You can easily integrate also these logs and metrics
[21:32.000 --> 21:34.000]  by visualizing them.
[21:34.000 --> 21:36.000]  So with it, that was a very easy demo.
[21:36.000 --> 21:38.000]  I think we can continue.
[21:38.000 --> 21:41.000]  If you're interested in that, you can play with other plugins
[21:41.000 --> 21:43.000]  straightforward.
[21:43.000 --> 21:48.000]  And now maybe we can switch to my presentation.
[21:48.000 --> 21:50.000]  And maybe if you have any questions,
[21:50.000 --> 21:53.000]  we can jump into question presentation
[21:53.000 --> 21:55.000]  because we are already running out of time.
[21:55.000 --> 21:58.000]  Here are some references I'm giving out to you.
[21:58.000 --> 22:00.000]  For you.
[22:00.000 --> 22:03.000]  Yep.
[22:03.000 --> 22:05.000]  Any questions?
[22:05.000 --> 22:06.000]  OK, thanks.
[22:06.000 --> 22:07.000]  Thanks a lot.
[22:07.000 --> 22:21.000]  So, I have one more t-shirt.
[22:21.000 --> 22:25.000]  Yeah, I have a question regarding the plugins.
[22:25.000 --> 22:30.000]  For instance, with traffic, you have to kind of connect it
[22:30.000 --> 22:33.000]  to pilot and download them as it starts.
[22:33.000 --> 22:36.000]  How do you develop custom plugins for API 6
[22:36.000 --> 22:40.000]  and how do you package them with the API gateway?
[22:40.000 --> 22:42.000]  Sorry, once again, last few sentences.
[22:42.000 --> 22:44.000]  There was a door is open.
[22:44.000 --> 22:45.000]  Sure.
[22:45.000 --> 22:47.000]  I was just wondering about plugin development,
[22:47.000 --> 22:51.000]  like how rich is the ecosystem for API 6.
[22:51.000 --> 22:54.000]  And if I were to develop a plugin,
[22:54.000 --> 22:58.000]  is it easy to package with the binary
[22:58.000 --> 23:01.000]  or do I have to connect to some kind of pilot
[23:01.000 --> 23:02.000]  and it downloads it?
[23:02.000 --> 23:03.000]  Yeah, very nice question.
[23:03.000 --> 23:04.000]  I love it.
[23:04.000 --> 23:05.000]  Very nice question.
[23:05.000 --> 23:07.000]  Yeah, it's very straightforward
[23:07.000 --> 23:11.000]  because we have now support for five different languages.
[23:11.000 --> 23:15.000]  If you are using this support program in English,
[23:15.000 --> 23:16.000]  we have plugin runners.
[23:16.000 --> 23:18.000]  You don't have to build anything.
[23:18.000 --> 23:20.000]  You can just run.
[23:20.000 --> 23:23.000]  Of course, you need to create some binaries for Java,
[23:23.000 --> 23:25.000]  let's say, Java file.
[23:25.000 --> 23:27.000]  And this Java file has a connection
[23:27.000 --> 23:29.000]  by using Unix socket files.
[23:29.000 --> 23:32.000]  It can communicate with Unix socket files
[23:32.000 --> 23:34.000]  and you can exchange some log data.
[23:34.000 --> 23:36.000]  I mean, the request data.
[23:36.000 --> 23:38.000]  I had another presentation also.
[23:38.000 --> 23:41.000]  Maybe I can share with you after the session
[23:41.000 --> 23:43.000]  about how to write in Java plugin,
[23:43.000 --> 23:46.000]  how to create plugin maybe in Python and so on.
[23:46.000 --> 23:48.000]  This is a t-shirt for you.
[23:48.000 --> 23:50.000]  So I will keep it.
[23:50.000 --> 23:51.000]  More questions?
[23:51.000 --> 23:53.000]  But I don't have a t-shirt.
[23:53.000 --> 23:54.000]  Hello.
[23:54.000 --> 23:57.000]  How do you compare API 6 to other API gateways
[23:57.000 --> 24:00.000]  like Gravity, Kong, or...
[24:00.000 --> 24:02.000]  Which API gateways again?
[24:02.000 --> 24:08.000]  How do you compare API 6 to other API gateways on the market?
[24:08.000 --> 24:11.000]  How do I compare or...
[24:11.000 --> 24:13.000]  What's its main selling point?
[24:13.000 --> 24:15.000]  Sorry, once again.
[24:15.000 --> 24:16.000]  What's its main selling point?
[24:16.000 --> 24:22.000]  Why is it better than other API gateways?
[24:22.000 --> 24:25.000]  Code coverage, do you mean?
[24:25.000 --> 24:29.000]  Compared to the others, I think it's a little bit...
[24:29.000 --> 24:32.000]  I think it would be a little bit hard to listen from that.
[24:32.000 --> 24:34.000]  I will come to...
[24:34.000 --> 24:37.000]  There's a lot of API gateways on the market.
[24:37.000 --> 24:38.000]  Yes, yes.
[24:38.000 --> 24:40.000]  Why is API 6...
[24:40.000 --> 24:41.000]  Oh, do you mean benchmarking?
[24:41.000 --> 24:43.000]  Ah, okay, I got it.
[24:43.000 --> 24:46.000]  I'm not, of course, trying to sell API 6, right?
[24:46.000 --> 24:47.000]  Even if I have a t-shirt.
[24:47.000 --> 24:48.000]  I'm just giving out a t-shirt.
[24:48.000 --> 24:50.000]  I'm an open source contributor
[24:50.000 --> 24:55.000]  because one reason is it provides hot reloading of plugins.
[24:55.000 --> 24:57.000]  You don't have to stop your instance.
[24:57.000 --> 24:59.000]  You don't have to stop the instance.
[24:59.000 --> 25:03.000]  You can just run these plugins on the fly.
[25:03.000 --> 25:05.000]  You can switch off one plugin.
[25:05.000 --> 25:09.000]  You can write your custom plugin enable while this API 6 instance is running.
[25:09.000 --> 25:13.000]  This is one of the advantages that any other API gateways they don't provide.
[25:13.000 --> 25:16.000]  And the performance is on the top now
[25:16.000 --> 25:18.000]  compared to com API gateway.
[25:18.000 --> 25:20.000]  Our performance is high, twice faster.
[25:20.000 --> 25:22.000]  You can check it out.
[25:22.000 --> 25:24.000]  And we have...
[25:24.000 --> 25:28.000]  API 6 not only has API gateway, it has also English controller.
[25:28.000 --> 25:31.000]  You can use it for Kubernetes English controller
[25:31.000 --> 25:34.000]  or even you can extend it to a service mesh
[25:34.000 --> 25:36.000]  for your Kubernetes services.
[25:36.000 --> 25:38.000]  That's one of the advantages.
[25:38.000 --> 25:41.000]  But these advantages also exist, of course.
[25:41.000 --> 25:43.000]  It is open source, too.
[25:43.000 --> 25:44.000]  Yeah.
[25:44.000 --> 25:46.000]  I hope I could answer.
[25:46.000 --> 25:48.000]  Sorry, it was a little bit hard to listen to.
[25:48.000 --> 25:50.000]  Okay.
[25:50.000 --> 25:56.000]  Could you please elaborate a little bit on the scalability of those plugins?
[25:56.000 --> 25:58.000]  It would be nice.
[25:58.000 --> 26:01.000]  You're worrying about how it's scalable, right?
[26:01.000 --> 26:04.000]  How much it can carry?
[26:04.000 --> 26:06.000]  How much observations can be done?
[26:06.000 --> 26:08.000]  Good question.
[26:08.000 --> 26:10.000]  You can enable multiple plugins.
[26:10.000 --> 26:14.000]  It's this up to you how many plugins you would like to use.
[26:14.000 --> 26:19.000]  And there should be some problems when it comes to millions,
[26:19.000 --> 26:21.000]  maybe billions of API calls.
[26:21.000 --> 26:24.000]  That's why we have different deployment modes.
[26:24.000 --> 26:28.000]  For example, you can deploy API 6 as a standalone
[26:28.000 --> 26:33.000]  or maybe you can deploy with multiple API 6 instance one storage.
[26:33.000 --> 26:37.000]  Or one API 6, different types of storage.
[26:37.000 --> 26:42.000]  Because ITCD is compared to other relational database,
[26:42.000 --> 26:45.000]  non-relational database, super fast to first data.
[26:45.000 --> 26:49.000]  That's why it's also easy to deploy in many places.
[26:49.000 --> 26:52.000]  That's scalable without any issue.
[26:52.000 --> 26:55.000]  Yeah.
[26:55.000 --> 26:58.000]  Does it make sense?
[26:58.000 --> 27:00.000]  Yeah, just a quick question.
[27:00.000 --> 27:04.000]  I saw there you had API 6 EngineX.
[27:04.000 --> 27:08.000]  Is EngineX the underlying API gateway?
[27:08.000 --> 27:12.000]  And API 6 sits on top of it and expands the capabilities of it.
[27:12.000 --> 27:15.000]  Because EngineX is quite limited in that.
[27:15.000 --> 27:16.000]  Yeah, true.
[27:16.000 --> 27:20.000]  EngineX is the root of the API 6.
[27:20.000 --> 27:23.000]  It's built on the top of EngineX.
[27:23.000 --> 27:26.000]  But it provides additional features, right?
[27:26.000 --> 27:30.000]  Yeah.
[27:30.000 --> 27:32.000]  We have one more question.
[27:32.000 --> 27:34.000]  I'll try to make it short.
[27:34.000 --> 27:39.000]  Can API 6 work with GraphQL APIs?
[27:39.000 --> 27:40.000]  Yes.
[27:40.000 --> 27:43.000]  GraphQL, I think nowadays became very popular also.
[27:43.000 --> 27:44.000]  Sorry.
[27:44.000 --> 27:47.000]  I think Nikola also had the talk about tracing, right?
[27:47.000 --> 27:51.000]  He has some of the talks about this GraphQL.
[27:51.000 --> 27:55.000]  We are improving, we're contributing more now,
[27:55.000 --> 27:57.000]  massively on the GraphQL.
[27:57.000 --> 28:00.000]  My answer is it's possible, yes.
[28:00.000 --> 28:03.000]  You can try it out.
[28:03.000 --> 28:06.000]  Any other questions?
[28:06.000 --> 28:07.000]  OK.
[28:07.000 --> 28:09.000]  Thanks a lot, everyone.
[28:09.000 --> 28:10.000]  Thank you for coming.
[28:10.000 --> 28:11.000]  Thank you.