Our world is leading towards technology and innovation. Businesses, organizations possess a large amount of sensitive data and hence google introduces a cloud computing service which is Google Compute Engine (GCE). Google Compute Engine (GCE), is the infrastructure service of Google Cloud Platform. Amazon EC2 was announced in 2006 while Microsoft added VMs to Azure in 2012. Google announced the general availability of GCE just in late 2013.
Rather than being the laggard in the IaaS segment, GCE had some more advantages. Administrators can choose a region and zone where specific data resources can be stored and used. Right now, there are 3 regions of GCE: the United States, Europe and Asia. Each region possess two availability zones and each zone supports one of these two- Ivy Bridge or Sandy Bridge processors. GCE offers a suite of tools for administrators to create advanced networks on the regional level. Here are five amazing features of Google Compute Engine (GCE). Let us see one by one.
Five secret features Of Google Compute Engine (GCE)-
1. Shared Storage-
To find NAS or SAN appliances deployed in enterprise data centers is very common. They offer shared storage to applications and end users. There is a significant large barrier between on-premises and cloud in the form of shared storage. To imitate NAS -like configurations in the public cloud becomes very difficult for Enterprise customers. Configuring NFS or other shared file systems will negatively impact the performance of applications. There are some block storage devices like Amazon EBS or Azure Page Blobs that can be attached to one at a time.
This restricts the functionality of the disks by limiting to just one VM. Azure and AWS offer shared file system services but still they don’t reach to the performance of a SSD-backed storage device. You can attach one persistent disk to multiple running instances in Google Compute Engine. The main reproach is- the disk is available in read-only mode to the VMs. When data is in read only mode, it helps customers to copy multiple scenarios close to on-premises deployment.
2. Disk Resizing-
Redirecting to the virtual infrastructure possesses many benefits. These benefits include rapid scale out and scale in. It is difficult to emulate the same with block storage when IaaS providers offer elastic scaling of compute resources. SSD-backed persistence for block storage devices are independent of the VM. The data stored in block storage will be available even after a smooth or abrupt termination of virtual machines. Amazon’s EBS (Elastic Block Storage) can be attached to EC2 instances. And also EBS volumes can be periodically backed up.
At the time when an EBS volume runs out of space, customers should stop the running instance, detach the EBS volume, restore the latest snapshot to new EBS with a larger volume and to reattach that instance before starting it. This includes downtime of the instance. Microsoft Azure is the same with respect to resizing an attached disk. Google Compute Engine has the Persistent disks which are the counterparts of Elastic Block Storage of Amazon EC2. It provides long-term, durable storage to VMs.
Just now, online disk resizing is introduced by google without any downtime of virtual machines. It avoids the bulky workflow of taking the VMs offline, restoring the snapshot, and also reattaching them. Hence you should gracefully drain the connections before starting the resizing task when you are running I/O-intensive workloads. To resize the live disk, the customer can either use the portal or the command line interface. After completing the disk space expanding, they must follow the routine specific to the operating system to claim the available space. Online disk resizing is an amazing feature of Google Compute Engine that no other competitors provide.
3. Sustained Usage Discounts-
Economies of scale in the IaaS business directed by the infrastructure use. It is just like the aviation industry in which operational cost is decreased with the increase in passengers. The IaaS providers invest into infrastructure capacity upfront hence they have to ensure the most ideal resource use. Cloud providers offer infrastructure at a low cost than the on-demand price to motivate the customers to run their workloads for a longer term. Example of such pricing scheme is- Amazon EC2 Reserved Instances feature. When a customer commits to a 12-month term services, Microsoft offers a 5% discount on Azure.
At the starting days, it was simple to understand the concept of reserved instances. But as the demand increases and new usage patterns come into focus Amazon made it complicated and confusing. For this, customer have to spend more time to calculate the reserved instance pricing than choosing the right EC2 instance type. Google Compute Engine’s sustained usage discounts reward clients for their use of compute resources. For this, there is no need to commit for the long-term to enjoy discounts.
All the customer should just simply continue running the VM for a whole month. At the end of the billing cycle, Google automatically adds the discount to the bill. The discount increases as the use getting up to a 30% net discount for instances that run the entire month. Regardless of whether the customer doesn’t run the same VM for the whole month, Google treats multiple, non-overlapping instances running a similar region and zone as one VM to apply the discount.
These instances are called inferred instances in GCE terminology. By offering discounted price to all the GCE customers, this feature fills the gap between on-demand pricing and reserved pricing. No other IaaS provider could match Google in offering sustained usage discounts.
4. Preemptible VMs-
Continuous discount use comes the preemptible VMs capability of GCE. It is designed for the same reason as continuing the discount use- to drive more use of foundation. Amazon launched Spot Instances in 2009. This allows customers to offer on unused EC2 instances. The cost on per hour basis is decided by Amazon and it changes according to the supply and demand for Spot instances. This model works like the spot market in the financial industry where financial instruments or commodities are traded for immediate delivery. Spot instances may be ended whenever the offer cost goes up.
It is expected that customers should run only those applications that can tolerate the abrupt shutdown. As per Google, preemptible VMs are more cost-effective, and also short-lived compute instances appropriate for batch jobs and fault-tolerant workloads. They may offer price reduction to 70% when compared to regular VMs. If we say that preemptible VMs and Amazon EC2 spot instances are the same, then what’s the difference? Not at all like EC2 spot instances, customers need not bid unused capacity. It avoids the pain that generally occurs with complex bidding and gambling on fluctuating market movements. Any Google Compute Engine VM can be launched in preemptive mode.
The VM might be ended under two conditions – first is- the particular region where the VM is deployed has run out of capacity, and the second is,- the VM has been running throughout the previous 24 hours. VMs that can be launched under the preemptive mode don’t have any limitations. Fault-tolerant workloads, for example, big data clusters, media transcoding, and web crawling are great candidates for preemptible VMs. While Amazon has EC2 spot instances, Azure doesn’t have the bidding mechanism to launch VMs.
5. Custom VM Sizes-
Migrating workloads to the cloud is a complex task and it has the biggest challenge such as selecting the appropriate instance types. Also it will be an overwhelming task to map on-premises resource configuration with virtual infrastructure exposed by cloud providers. Amazon EC2 continued including new instance types and families to its portfolio. The list confusing and overwhelming because of more than 40 instances classified into six families. Azure is simple to choose the VM types. It has a list of more than 29 instance types spread across 4 categories and this list is continuously growing. Cloud migration and deployment comes has a significant cost due to the complexity of choosing the correct instance type. Customer can’t change the instance type when switching to a reserved instance on Amazon EC2. The only alternative is to resell the reserved instance in the marketplace.
Google Compute Engine offers custom VM sizes where customers can precisely choose the number of CPU cores and the amount of memory required for their workload. Customers can choose anywhere between 1 core to 32 cores of CPU and up to 6.5GB of RAM per vCPU according to the zone where the VM is launched. The VMs can run Linux or Windows operating systems. When combined with sustained usage discounts and preemptible VMs this feature offers great value to customers.
These are some features used by customers to migrate large content management systems to Google Compute Engine.
Are you looking to migrate your business data to google cloud? But confused to decide whether to choose or not? Consult with solace experts for appropriate solutions. We have dedicated experts to help you through consultation and development with Google compute engine. Connect with Solace and get a free quote for software development with new technologies. We will be happy to help you.