The data center engine can ingest real-time data streams from the IoT, data-driven customer interactions and business insight platforms.
Startup Mesosphere, which developed the first open-source-based data center operating system only during the past 18 months, already has expanded the scope of its young system.
The San Francisco-based company on Aug. 21 introduced Mesosphere Infinity, the first real-time, enterprise-ready open-source data center engine that can ingest at high scale streams of data from the Internet of things (IoT), real-time data-driven customer interaction solutions and real-time business insight platforms. All three of these factors are key drivers in the design and building of new-gen data centers.
Developed in collaboration with Cisco Systems, Mesosphere Infinity runs on containers and is a turnkey, full-stack offering optimized for big data and IoT workloads. It is powered by the Mesosphere Data Center Operating System, or DCOS, which is built around Apache Mesos and an integrated, commercially supported stack of Apache Spark, Apache Kafka, Apache Cassandra and Akka.
Components in Infinity include:
A high-performance publish/subscribe message bus designed for high availability and data durability, with minimal latency. Kafka acts as a central data backbone and enables loose coupling between applications (Akka), data processing (Spark) and data persistence (Cassandra) services.
An event-driven middleware framework, for building high performance and reliable distributed applications in Java and Scala. Akka decouples business logic from low-level mechanisms such as threads, locks and non-blocking IO. Akka provides a set of high-level abstractions (including Actors, Streams and Futures) and simplifies development of IoT and other message-driven applications.
: A fast and general-purpose engine for large-scale data processing in memory. Spark supports streaming analytics, machine learning and graph computation. Spark's in-memory primitives provide performance up to 100 times faster than that of traditional open-source big data frameworks.
A database that provides for high scalability, high availability and linear scalability, making it the ideal platform for mission-critical data. Cassandra support for replicating data across multiple data centers is best-in-class and enables low latency geographically distributed applications and peace of mind of knowing that you can survive regional outages.
Mesosphere Infinity supports a broad range of applications, including:
Harnesses the power of billions of connected devices, sensors and other data sources to create groundbreaking new products, disrupt existing business models or more efficiently manage your supply chain.
Detect in real-time problems such as financial fraud, a structural defect, a medical condition and other errors.
Analyze real-time and historical data to forecast probabilities in order to better understand customers, products and partners and to identify potential risks and business opportunities.
Enterprises can deliver a unique experience in real time that is relevant and engaging based on a deep understanding of the customer and current context.
Competes With OpenStack, to an Extent
Infinity is, in many ways, a competitor to OpenStack, an open-source enterprise system developed by NASA and Rackspace several years ago.
By using Infinity, Mesosphere said, enterprises can inherently capture and analyze new business opportunities created by the rise of ubiquitous data—including fraud and anomaly detection, real-time personalization, and IoT applications, such as connected consumer devices and supply chain optimization.
Mesos, the kernel of the Mesosphere DCOS, is a 6-year-old Apache open-source project, conceived at the University of California, Berkeley, that was announced as a joint collaboration with Mesosphere at DockerCon EU in December 2014. The company has come a long way in the nine months since then, as more and more enterprises retool their data centers to run DCOS.
Mesosphere DCOS is a highly scalable engine that enables the running of services and applications across a cluster of machines in a data center or cloud. It is highly container-driven. It combines the Apache Mesos cluster manager with a number of open-source and proprietary components and allows services to be deployed and managed through both a custom Web UI and command-line interface.
No other operating system available at this time can run to the scale that Mesos can, Matt Trifiro, Mesosphere senior vice president of marketing, told eWEEK
Cisco Systems an Important Partner in Distribution
Cisco Systems, as a key collaborator with Mesosphere on the development of Infinity, will be one of the first distributors to enterprise customers. As part of its development contributions, Cisco has optimized Cisco Intercloud Services for Mesosphere Infinity and already has a number of customers running the stack in production.
Intel has validated the stack as being suitable at scale on Intel hardware for the most demanding production deployment requirements, Mesosphere said.
For more information, go here.