Welcome back. In this lesson, we'll be talking about metadata. We'll talk about the metadata server and how to get and set metadata.

Let's start with a quick explanation of what metadata is in this context. Metadata is a set of key value pairs containing information related to running instances or our project. This can be stuff such as project-level settings. Maybe things like API keys for external services. And instance-level information are things such as tags or the zone that we're running in or the hostname or whether the instance is pre-emptible or not.

So, as I mentioned, there are two types of metadata: instance and project. And they're accessed through the same base URL. Now we have three URL options and they're all going to direct you to the same place. There's the full URL, that's metadata.google.internal/computeMetadata/v1. There's the short URL which is metadata/computeMetadata/v1. And then the final option is to use the IP address. And then, you'll either append project or instance to the end of the URL to access the metadata specific to those topics.

Let's look at how to actually get and set data. Like most everything else, we can use the API, the SDK or the console to get and set metadata. However, to get metadata from a compute engine instance, using these options would add extra steps that we just don't need. From a running instance, all we need to do is call the URLs and add these specific values to the end of the URL. You could do it in code, or you could use a tool such as curl or Wget.

Let's connect in to an instance that we have running and see how to use these. We're connected via SSH to a compute engine instance. So, we're going to use cURL and we can start getting data from the metadata server. If we start with the base URL, we have two options here. You can see we have instance and project. And these values are what we use to append to the end of the URL and this allows us to drill-down in to the different keys. So, if we append instance to the end of this, you can see we have some additional things that we can drill in to here.

Let's check out tags. So, all we have to do is add tags to the end of the URL and you can see we get some tags back here. These ones allow us to have http and https traffic through the firewall. We can query for things such as the zone that we're in. And, in this case, we're in east1-d. And if we edit the URL, we can query for project-wide settings as well. We can see here that we have just two attributes. One is a Google-supplied and the other is a value that I've added. So, if we want to read the value for the dotnet environment, we'd append that string. And we can see that we have a value there.

Now, to change it, if we jump in to the console on the compute engine page, in to the metadata section, we can change that value. Let's change it to staging. Okay, and then we'll save it. Great, now, if we jump back in to the command line, we can refetch that and see if it says staging. So, there it is. It's changed to staging. It's a project-wide setting.

So, metadata will make our script more dynamic, because we don't need to hard-code any values. And, we don't need to know, at the time that we write our code, some of these values. For example, we don't need to know the project ID because we can fetch it dynamically. And it's not uncommon to use environment variables to store things such as API keys, connection strings, access tokens, etc. So, if we use project metadata for those sort of things, we can ensure that instances will have access to the data they need and have it in a centralized location for us to easily manage. Okay, that's gonna wrap up this lesson.

In our next lesson, we'll talk about startup and shutdown scripts. So, if you're ready to keep going, then let's get started.