From the course: Google Cloud Platform Essential Training for Administrators

Compute service selection - Google Cloud Platform Tutorial

From the course: Google Cloud Platform Essential Training for Administrators

Start my 1-month free trial Buy for my team

Compute service selection

“

- GCP Compute Architecture Concepts include many choices, similar to what we saw with violin data choices and I've pulled the ones that I use most frequently here and really be aware. This is a subset and really just an introduction to the topic. So let's look at what we're going to be reviewing in this section. For virtual machines, we have GCE or Google Compute Engine and some of the considerations there are the complexity of both configuration and scaling on VMs. So increasingly, I have customers turning to what's called a serverless solution, which is the second thing listed here or called Functions. These are Google Cloud Functions, and they're analogous to for example, Amazon Lambda, or Azure Functions, in many ways. You need to verify your particular programming language support because although Google is increasing the programming languages supported to include the most popular languages such as Python, if you have a language that's in less usage, say Haskell, you would have to use a different solution. The next choice category that we have is managed container clusters and the two big service offerings here are Cloud Run, which is one of my favorite GCP services. You'll see when we dig into it, and then the older but still around Google App Engine or GAE. I use Cloud Run for container testing and container optimization. So it's just really an essential tool in my toolkit. GAE, I find out there in the wild, I'm not really developing actively with it but it really is a great way to get simple websites up very quickly. That's what it was originally designed for. So sometimes I see it, startups, hackathons, that kind of stuff. I'm sure it's out there in production, just not seeing it as much. So I'm not actually going to include much information about it other than to mention it here and then at a very, very complex level but powerful is the ability to manage raw container clusters using the powerful Google Kubernetes engine or GKE Service. That is such a complex service and it's one that I see customers who really shouldn't be using it. And so I'm actually going to put that into my enterprise GCP course in the library. I'm going to move it out of here because really you need to use the right tool for the job. And for most of the beginning architectures, smaller applications, VMs or Functions or Cloud Run will work just fine. And I see customers giving themselves in a lot of trouble by not understanding what the overhead is. I mean, yes, there are scenarios you're going to need a GKE but you're also going to need a team of engineers to manage it. Also, I'm going to showcase a tip that I use which is to use the Google marketplace that we saw in the previous section for data sets but this time we're going to use it for pre-built images and solutions. We're going to do that with WordPress.

Contents