Review: In this four part blog series, I am focusing on the breadth and ease of use of four leading cloud platforms: AWS, Google Cloud Platform, Azure and vCloud Air. In Part 1, I focused on the market in general and Amazon AWS. For Part 2, we will take a look at Google Cloud Platform.
The steps I have chosen for comparison of IaaS providers is:
- Create a Wndows machine from the service catalog
- Start the Windows machine
- Open a remote desktop to the Windows machine
- Display performance of Windows machine using IaaS tools
Create Windows machine from service catalogLogin to the Google Cloud Platform console
There are a good number of services to choose from in the Google Cloud Platform, but not as many services as Amazon offers.
Step 1: Click "Compute Engine" under Compute
Step 2: Click "Create Instance" button
Step 3: Name, size and choose the OS for your VM. You can even choose SSD or standard backing disk. I will choose Windows Server 2012 for the OS and standard persistent disk. Seeing the price change as you choose features is very helpful.
Here are all the choices for your OS image. You can, of course, create your own image.
Step 4: Choose management and automation options
Step 5: Choose disk options
Step 6: Choose networking options
Step 7: Choose SSH options if you want a unique key just for this VM
Step 8: Click Create. You can also examine what the REST or CLI command would be to automate this creation.
Within a few seconds, your VM will be created and ready.
Step 9: Set a username and password for your VM
Step 10: Open a console. Either download the RDP file to use with your RDP client or click RDP to open the console in the Chrome RDP client.
Voila! Windows Desktop.
Now, on to performance monitoring using cloud provider tools
Step 11: Choose VM Instances from the Compute Engine menu. A CPU utilization graph will display, by default, for all of your VM instances. Clicking on the name of one of your instances will bring up instance-specific graphs. You can switch to graphs for disk throughput, disk IOPS, network throughput or network packets.
The graph options are rather limited compared to the graphs in AWS where you can choose metrics to place on a graph, but they suffice for a general overview of utilization by your instances. One could use built in operating system monitoring such as perfmon or sar or vmstat for more detailed metrics.
In general, I found Google Cloud Platform very ease to use with fewer steps than AWS, even if the built-in monitoring is limited.
Now on to Azure for the next post. I welcome your feedback on this and future posts.