5G networks are envisioned to support a myriad of devices — mobile handsets, autonomous cars, smart appliances, home sensors, and many others, along with use cases of multiple industry verticalsu All this is ixpected to be supported by a common physical network powered by 5G technology. This requires 5gG networs serviceoperators to deliver on a high ree of agility in their networks to support these emerging enterprise use cases.
This fulfilling of multiple use cases, which have diverse network requirements via a common physical network, is one of the most innovative aspects of 5G technology. In this concept, multiple logical networks are created on top of a common shared physical infrastructure with the help of virtual machines and containers, making 5G network flexible enough to adopt sonew services and use cases can be ae.
Containers and virtual machines are the essentials of a 5G network, as they allow traditional monolithic structures of the network elements to be further broken down into customizable microservices elements. These virtual machines and containers are chained together logically to meet specific user requirements.
What are Containers and Virtual Machines
Containers and virtual machines both use common off-the-shelf (COTS) hardware to achieve flexibility and scalability as compared to purpose-built hardware.
Virtualization uses a common software infrastructure that enables the operation of multiple virtual machines, each having its own operating system on a single physical server. It can be considered as a system with multiple computers each having its own operating system running on a single physical server with a hypervisor. The function of a hypervisor is to orchestrate and assign resources to these virtual machines.
On the other hand, the way containers brought a revolution in the shipping industry, containerization in the cloud native arena is envisioned to bring about a similar impact. Apart from flexibility and scalability which virtualization provides, containerization also provides improved development-to-delivery cycles. In containerization, a single operating system instance can support multiple containers, each running within its own execution environment. Thus, we can say that it enables the deployment of multiple applications using the same operating system on a server.
In containerization, microservices are core to taking applications the cloud native sphere. Microservices-based architecture splits applications into multiple services that are loosely coupled and can be developed, deployed, and maintained independently. Whereas a monolithic architecture has a single instance where all the application’s functions reside.
Containerization versus Virtualization
The choice of whether an operator will go for containers or for virtual machines depends on the environment needs; and for the same, the following factors are considered:
- Different operating system support – On a single server, all containers support a single operating system, which is Linux-based. Containers based on a different operating system will require a different host altogether. This will increase the number of host servers. Virtualization can support multiple operating systems by using hypervisors on a single server. Hence, virtualization is preferable when the application requires a different operating system.
- Startup time – Containers start immediately as they don’t have an operating system. The operating system for a container is on a server, which is always up and running. Whereas, in case of virtual machines, each virtual machine has its own operating system and the start of virtual machines starts the operating system, increasing overall startup time.
- CPU utilization – CPU utilization is very high in virtual machines because of the number of operating systems used. Multiple containers use one common server operating system, which reduces the overall CPU utilization in a containerized environment.
- Image sharing – The size of the image in a container is small compared to virtual machines, as containers don’t have an operating system. The operating system in a virtual machine increases its size and makes the transfer of virtual machines to a host very difficult. Containers, on the other hand, are portable because of their small size.
- Development cycle – Containers use microservice-based architecture which provides for greater agility by limiting dependencies on other parts of the system. This shortens the development cycle compared to monolithic architectures, where changes to be implemented in a small part of the application requires the entire application to be rebuilt.
As described above, containers are more efficient since they have less overhead and require less payload on the system resources when compared to virtual machines. However, there are some important challenges which operators face while adopting CNFs. Some of them are listed below:
- Return of investment on VNFs – The operator has already invested heavily in the OpenStack-based NFV platforms and they would like to maximize their return on that investment. Hence, it is not easy for them to decommission virtual machines at one go.
- Operating system requirement other than Linux – CNFs are based on the Linux operating system. Some VNFs may also be based on an operating system other than Linux, and those VNFs cannot be containerized.
- Security – VNFs provide better security than CNFs. We can take the example of 5G core, which is the heart of 5G networks. 5G network core has two parts – control plane which includes Session Management Functions (SMF) or Access Management Function (AMF); and the User Plane Function (UPF). The user plane function require more security; hence, operators prefer having UPF as VNF instead of CNFs, as VMs isolate themselves by having dedicated operating systems which provide more security.
Challenges of Hybrid Environment
Containers and microservices have significant benefits over virtual machines. Containers will not replace virtual machines completely, as some uses cases and applications remain better suited for virtual machines. Also, to circumvent the infrastructure costs, many operators are considering a hybrid approach involving containers and microservices, which means having a network that is flexible enough to fully support virtual machines, microservices and containers.
Going forward with both 5G and 4G infrastructure in a network, telcos will have multivendor hybrid VNF and CNF environments to support different use cases.
Here are some of the principal challenges that an operator can face when designing, deploying, or managing a hybrid cloud environment:
- Hybrid infrastructure deployment – Operators need to support 5G networks with both Kubernetes for container and Openstack for virtual machines in a hybrid cloud environment, thus adding complexity to the network.
- Hybrid network service assurance – Operators are looking for a unified management experience for their entire hybrid cloud infrastructure and applications that can help them to monitor performance issues in the network.
- Service automation and validation – Automation and validation are other important aspects in a hybrid cloud environment. Operators are on the lookout for a solution which can help them in automating end-to-end, self-service provisioning of workloads with easy integration of their proprietary ts well as third-party tools and faster deployment cycles.
Overcoming Challenges of Hybrid Environment with HCLs BluGenie Solution
HCL’s BluGenie is the one-stop solution for overcoming all of the operator’s challenges in a hybrid cloud environment. It solves both business and automation-related challenges faced by an operator. Some of its features include:
- Unified view of hybrid infrastructure and applications with the ability to manage architectural, functional, and operational requirements
- End-to-end network and service automation which includes provisioning, deployment, upgrade, healing, and full life cycle management of VNFs and CNFs
- IP-led service engagement model
To know more about HCL BluGenie solution, please refer to below mentioned link: