Kubernetes has quickly become the most popular container orchestration solution with DevOps and Site Reliability Engineers. Kubernetes was developed and then open sourced by Google, it supports not just Docker (the defacto standard for building and running containerized apps) but also other container platforms. Containerization provides modern app developers with a means to continuously iterate and quickly deploy software updates multiple times a day in a fast, distributed, and isolated environment. Containers are going to be especially important to build and scale 5G apps.
As the telecommunications industry starts transitioning to 5G, it is preparing for a truly seismic shift. 5G is nothing like the telco industry has experienced ever before; with speeds of up to 100 gigabits/second, it will be up to 1000 times faster than 4G, opening up opportunities which were just not feasible before. 5G powered communication will enable even more apps to be real-time, and connect not just people but also billions of sensors to turbo charge the adoption of IoT.
The 5G network architecture also brings an tremendous business opportunity to telcos in the form of network slicing. Through network slicing, telcos can deliver customized apps and microservices that cater to each unique customers architecture and need. These new real-time apps and services will consume very high bandwidths, and will require economies of scale; i.e. cloud computing.
Along with tremendous potential, 5G brings with it a host of challenges, some of which telcos have never experienced before:
Real-time state-free apps and microservices in the cloud
As mobile consumers roam and switch from one network to another, device data and state change constantly. The state data needs to be collected, stored, and used in different aspects of the network so that the apps and microservices can access it instantaneously.The performance of commodity hardware in the cloud is much lower than what is typically desired by databases. Traditional databases already hampered by their legacy architecture are constrained even further by cloud hardware, and come no where close to delivering 5G’s demanding latency requirements.
Software defined networking (SDN)
5G comes with many new advanced features. Provisioning these features and managing network operations takes an herculean effort. Additionally, network function virtualization will be required to support network slicing. SDN is a necessity to deliver all this new complex functionality.
Unlike technology vendors, traditionally telcos had not been in the business of iterating, developing, and deploying apps and microservices. With the new 5G enabled business models, telcos have a lot of catching up to do and will need to quickly adapt the continuous integration and continuous delivery (CI/CD) methodology and transform them self into (at least partially) becoming a software vendor.
Since inception VoltDB has been has been the gold standard operational database for telecom software solution providers and operators. Recent enhancements have made VoltDB even faster, easier to use and deploy, these improvements were made specifically with 5G in mind. Here’s why VoltDB is the perfect choice for 5G:
- Unlike streaming solutions and some databases; VoltDB is a operational database that stores state data. VoltDB also has always prioritized ACID, in fact it offers the strongest ACID guarantee on the market: strict serialization. VoltDB’s 100% ACID makes it the ideal choice for transactional analytics on CDR and XDR data. With built-in machine learning VoltDB can make complex decisions in real-time for customer offer management, retention, fraud detection, and a host of other telco use cases.
5G’s latency requirements are causing telcos to move to active-active cross data center replication (XDCR), as the speed of light makes a single large data center too slow for end users. VoltDB is a perfect fit cloud based 5G apps and microservices; it’s in-memory distributed architecture was designed to run on commodity hardware and since inception it has offered a telco grade low, long-tail latency, along with active-active XDCR.
- VoltDB’s stored state data and strict serialization can help take some of the complexity out of SDN by maintaining state and ensuring 100% ACID on the network operations data. This enables the app/microservices developer to not worry about programming for ACID, saving valuable time and thus bringing the app/microservice to the market faster.
- VoltDB is architected to run in an distributed containerized environment, a lot of customers have been running VoltDB Docker containers for years. And now VoltDB supports Kubernetes, allowing it to seamlessly fit into CI/CD pipelines. As network slicing delivers new apps and microservices, container orchestration with Kubernetes turns many tedious and complex tasks into something as simple as a declarative config file, allowing for continuous and frequent deployment. Kubernetes is designed for stateless web apps that can easily spin up new interchangeable instances in case of node failure or scale-out. Up until now, orchestrating SQL databases in Kubernetes had been very challenging as operational database systems have state and can’t just be spun up or spun down on a moment’s notice.
VoltDB’s support for Kubernetes enables simplified:
- Spin-up and spin-down of VoltDB docker container clusters
- Monitoring of running clusters
- Cluster management and updates