Even in multinational companies, it’s an unusual scenario to have all the departments be in a hurry and running on deadlines all the time. A successful company has its own work rhythm. This may mean one or two departments require different business tools like a cloud computing resource to execute high-risk and urgent tasks. At the same time, another one or two departments are waiting for the pressing tasks to be done so they can execute the next steps. This process is executed steadily up until the work is done. With this rhythm, you can see that it’s not all the time that a department needs a cloud computing service to finish a project. For these kinds of instances, Google has presented a cheaper way to use the cloud for low-priority tasks.
What is Preemptible Virtual Machine?
The Preemptible Virtual Machine is a cloud platform introduced by Google just a couple of months ago. It is a service that enables computing resources at a lower cost, which is perfect for tasks that aren’t flagged as “urgent” or “hot,” and are thus considered interruptible. Google has computed that the new cloud platform can help a company save up to 70 percent compared to the pre-set Computer Engine setting. It is a cheap option, as budget-friendly as $0.01 per instance per hour. This is significantly lower compared to the majority of inexpensive Virtual Machines that charge between $0.03 and $0.11 on an hourly basis.
What is the catch?
The cheaper option can easily lure companies that use cloud services, no doubt. But one factor to take into consideration with the Preemptible VM is its catch – the machine may just abruptly stop working even when you’re actively using it. It may also stop with close to nothing of a warning when it reaches its peak periods.
What is Google saying about this drawback?
Google doesn’t see it as a particular cloud service drawback. In fact, the company points out that the service (in beta) is suitable for productivity tasks that are tolerant to faults, and that are easy to distribute to various Virtual Machines. Even critical workloads, such as data analysis, genomics, modeling, and simulation, which require computational prowess, can be run intermittently using the VMs.
What processes are implemented by Google to offer this cloud service?
To present this cloud computing resource to the buying public, the search engine giant will make use of the free capacity right in its data center. When there is a demand on these spaces and Google requires more cloud resources, then the virtual machines connected with the Preemptible Virtual Machine are brought back. This is where the interruption on the current processing happens. As a warning, Google will flash a warning with a 30-second window to save that, by computation, should be enough to save the interrupted task. The upside to this is that the VM can go on working straight for more than a day.
This new cloud service feature of Google is somewhat a direct completion with the leaders in cloud computing, Amazon Web Services and Microsoft.
What are your thoughts about the Preemptible Virtual Machine of Google? Share your two cents below.