Kong Kong Konnect Tutorials, 28 Sep 2021

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Microservices Monitoring and Logging with Kong Konnect

Description

It's quick and easy to set up microservices monitoring and logging with Kong Konnect. Try it for free: https://bit.ly/3A9njxz

▬▬▬▬▬▬ TIMECODES ▬▬▬▬▬▬

0:00 Intro and Starting Setup
1:24 Set Up Kong Konnect
5:06 Add Services to Kong Konnect
6:50 Add Monitoring to Kong Konnect with Prometheus Plugin
7:47 Configuring Monitoring in Kong Konnect with Prometheus and Grafana
9:28 Adding Logging to Kong Konnect with the Loggly Plugin
10:33 Conclusion

Follow along in the blog version: https://bit.ly/3o4fQfG

#Prometheus #Grafana #Kubernetes #APM #Microservices #Loggly #APIgateway #Kong

A

In this video we're going to look at how easy it is to set up services on con connect and then walk through how we'll hook our services into monitoring and logging solutions quickly and simply with kong plugins.

A

As an introduction, we already have three api services deployed to gcp cloud functions. These are simple: api servers, a user's api, a products api and an orders. Api each service has two get endpoints list, all in a get details: endpoint and one post endpoint for creating a resource just to demonstrate our users api. Briefly, we can send a get request with curl to get all users.

A

And there we see the ids and names of the users returned by our api.

A

We can send a post request with a json data payload containing the name of a user, and this creates a user.

A

If we send a get request to that endpoint with a user's id as a path parameter, we fetch the details for that single user.

A

The products and orders apis work. Similarly, next, let's get set up with kong connect.

A

First, we've created a kong connect account and we're logged in conconnect requires that you have a kong gateway, runtime deployed when your con gateway runtime starts up. It reaches out to conconnect establishing a connection for updates.

A

We've installed, our con gateway runtime on our local machine, we're using kong gateway enterprise in free mode at version 2.4.1.1.

A

We need to configure kong gateway with certificates for authentication to con connect. We click on run times and then configure runtime.

A

Since we have our runtime deployed to our local linux machine, we click on linux, then we click on generate certificate, we'll want to copy each of these three strings to their own files, on your local machine in your etc. Cong folder create a subfolder called certificates.

A

We'll copy the cluster certificate string to a file called tls.crt.

A

The certificate key string will copy to a file called tls.key.

A

And the root ca certificate we will copy to a file called ca.crt.

A

Next, we need to configure our kong gateway so that it starts up ready to reach out to con connect on our con connect. Runtime configuration page. We see a list of configuration parameters copy those to your clipboard.

A

Our local kong installation comes with a kong.conf.default file. We copy that file to a new file called kong.conf.

A

That's the configuration file location that kong will look to when it starts up.

A

We open that file and at the bottom we paste in the configuration parameters we just copied from con connect for the last three lines of the parameters we'll need to enter the absolute path to the three certificate related files that we just created.

A

So, for example, cluster cert should point to etc cong certificates, slash, tls.crt, cluster cert key, should point to etc cong certificates, slash, tls, dot, key and lua ssl trusted certificate should point to etc certificates, slash ca, dot, crt.

A

We also have one other configuration that we want to modify at around line 517.

A

We want to set the status, listen parameter so that kong's status api will be exposed for a monitoring tool like prometheus, which is what we'll be using to receive metrics. This parameter should be set to 0.0.0.0.

A

Colon for port 8001.

A

With our parameters in place, we start up kong on our local machine.

A

Back at con connect in our runtime manager, we refresh the page.

A

And there we see our locally deployed con gateway. Runtime has connected with conconnect.

A

Now we can begin adding our services in the service hub. We click on add new service. Let's add the users service. First, we set a name version and description for our service. After creating our service, we navigate to our 1.0 version to create a new implementation.

A

Here we add the url for our upstream service. We paste in the url for our gcp cloud function where our users api, resides. We click on next to configure the route for our service.

A

We set a name for our route, choose the http protocol, since we are running kong gateway locally and then enter the methods to listen for, along with a path that our local kong proxy should listen on when we click on create our service updates. With this new implementation, conconnect reaches out to our con gateway, runtime and updates it with our new service implementation.

A

Now, when we send a curl request to localhost 8000, that's our kong proxy, slash users. We see our json response with our users just to save us some time in this video I've gone ahead and performed the exact same steps to add our orders: api service and our products api service to con connect.

A

Now we can send a curl request to localhost 8000 orders and see our list of orders, and we can send a curl request to localhost 8000, slash products and see our list of products.

A

Next, let's add in monitoring in conconnect. This is quick and simple for each of our services. We navigate to our latest version at the bottom. We click on new plugin.

A

We search for the prometheus plugin when we found it. We click on enable we can add a tag to our plugin for this service, just to differentiate its metrics from other services also using prometheus, and we quickly do the same thing for our order: service and our products service.

A

Next, we want to update our prometheus configuration, so it knows where to look for metrics. We open up the configuration yaml file and we set the targets in the scrape configs to look to localhost 8001.

A

And we set our scrape interval to 5 seconds.

A

We start up our prometheus server in one terminal window and we start up our grafana server in another terminal window.

A

In our browser we pull up grafana and we log in.

A

First, we will need to add our data source, we'll add a data source called prometheus, which is located at localhost port 9090.

A

After we've added prometheus as a data source, we want to add a dashboard for some good visualizations. Fortunately, kong has an official dashboard for grafana in grafana. We click on the side nav bar to import a new dashboard.

A

Kong's official dashboard has an id of 7424.

A

We enter it, find the dashboard, select our prometheus data source and then click on import and with that we're able to start seeing some visualizations of the monitoring metrics that have come through our services.

A

Next, we're going to add logging to our services by enabling the logley plugin. This plugin will send log messages from our services to a centralized log management solution, similar to our enabling of the prometheus plugin. We go to each service, click on the version and click on new plugin.

A

We search for the logley plugin, and then we enable it. We can add a tag that differentiates this particular service and for config key. We make sure to enter the customer token associated with our logley account.

A

We perform the same steps for our other two services with that our logley plugin setup is complete.

A

In the background we made a few additional curl requests to our services just to get some activity there and then back in logley. We start to see some entries from our recent request to our services.

A

In this demo, we walk through the setup of kong, connect configuring, a connection with a locally deployed kong gateway runtime, we added services to conconnect, and then we showed how to enable and configure a monitoring solution with the prometheus plugin and a logging solution with the logley plugin.

A

Thank you for joining us.