Vapor IO Empowers the Autonomous Edge with Synse 3.0

By AIT News Desk On Apr 18, 2020

Open source platform helps accelerate IT/OT convergence, making environmental data actionable within a data center

Vapor IO, creators of the Kinetic Edge platform, the first nationwide platform for edge colocation, networking and exchange services, announced Synse 3.0, the next generation of its open source API for making environmental data actionable within the network and data center. The developers of Synse 3.0 rebuilt portions of the code, making it massively scalable and adaptable, and now includes the 2.0 version of the SDK, making it extremely easy for third parties to build drivers that integrate new types of equipment and sensors.

“Synse 3.0 is a potential game-changer for how Section delivers edge workloads for our partners, and we’re looking to incorporate its environmental data into our patent-pending Adaptive Edge Engine”

“As thousands of geographically distributed, lights-out data centers emerge at the edge, we have no choice but to make our infrastructure and applications autonomous,” said Cole Crawford, founder and CEO of Vapor IO. “Without the IT/OT convergence enabled by Synse, you don’t have enough context to properly automate edge environments. Synse 3.0 provides an open source mechanism for exposing critical environmental information that impacts servers and workloads, further empowering the autonomous edge.”

IT/OT Convergence

In traditional data centers, the IT and OT rarely converge. The teams responsible for managing the servers and applications seldom have access to the OT performance data, which makes it difficult or impossible for them to make fully-informed decisions about resilience and workload optimization. For example, it’s much harder to diagnose or predict the failure of a hard drive without combining the drive’s performance over time with information from the surrounding facilities environment, such fluctuations in the HVAC system or air pressure at the server backplane.

Synse provides access to a wide variety of data streams, including the operating statuses of HVAC systems, fire suppression systems, security systems and sensors for things like pressure, vibration, temperature and humidity. Synse connects to environmental sensors and operational devices via plug-in drivers. Some common drivers, such as those for SNMP, IPMI, Intel AMT and Modbus, are already available in the Synse open source. Any drivers that don’t exist can be built with the Synse SDK.

The Synse Server organizes the data it consumes, then presents it through an API that can be queried with an HTTP request. Reliability and operations teams can then use their favorite tools to connect to a Synse server and ingest OT data for analysis and action. Synse data can be routed to open source analytics systems, such as Prometheus and Grafana, or be processed by proprietary systems according to user needs.

Workload Orchestration at the Edge

Edge environments demand the intelligent orchestration of workloads to make the best use of resources and to deliver high SLAs, such as those for resilience and latency. Software orchestrators, such as those used with container management systems like Mesos and Kubernetes, can be customized to include data from Synse when making workload placement and migration decisions. For example, a machine-learning algorithm could be built to predict server failures and that information could be used by an orchestration engine to start workloads on other servers, even those in other facilities, in order to mitigate any downtime caused by a failure.

“Synse 3.0 is a potential game-changer for how Section delivers edge workloads for our partners, and we’re looking to incorporate its environmental data into our patent-pending Adaptive Edge Engine,” said Wesley Reisz, VP of Technology at Section, a partner of Vapor IO. “The Adaptive Edge Engine allows our platform to dynamically reshape points of presence – and the workloads that run on them – based on real-time data. The possibility of extending those insights with Synse 3.0, with the added benefit of tying into tools that we already use like Prometheus and Grafana, could have an important impact on the way we run applications at the edge.”

Using Synse

Anybody can use, extend and adapt Synse free of charge. Synse 3.0 is licensed under an open source GNU General Public License v3.0, and includes the following components:

Synse API: Uniform API for presenting sensor and environmental information from data centers.
Synse Server: HTTP server that runs locally in a facility and which connects to operational equipment and sensors via plugins.
Synse Plugins: Device and protocol drivers used by the Synse server. A growing library of plugins are being made available as part of the Synse open source and developers can use the Synse SDK to customize existing plugins or build new ones.
Synse CLI: Tool for querying Synse from the command line.
Synse SDK: Software development kit for building Synse plugins.

For detailed information on getting started with Synse 3.0, read the documentation or access the GitHub repo.

Use Cases

Here are a few use cases for Synse:

Data center monitoring: Synse is designed to work alongside and augment existing monitoring and DCIM tools. It provides powerful data feeds capable of augmenting existing monitoring solutions. As an example, Vapor IO uses Synse to power the company’s NOC (Network Operations Center) as well as its Kinetic Edge Portal (which lets customers view environmental status surrounding their equipment).
Resilience: Service uptime in highly-distributed edge environments demands software-based highly-available (HA) systems, such as those offered by platforms like Mesos and Kubernetes. Synse can provide crucial environmental information that can inform automated schedulers, allowing them to make the best placement decisions to ensure software-based failover in the event of a facilities outage. For example, Vapor IO’s Kinetic Edge lets application providers treat an entire city as one large, virtual data center. By using technologies such as Kubernetes and Mesos to implement high-availability services on the Kinetic Edge using Synse, applications can achieve uptimes well in excess of the industry standard 99.999%.
Predictive maintenance & self-remediation: By using Synse environmental data to contextualize other data sources, more accurate and sophisticated machine learning models can be built that provide predictive maintenance and autonomous remediation solutions. For example, prior to Synse you might know that a hard drive failed; but, with Synse, you might be able to correlate those failures to power surges or temperature fluctuations that can be used to improve predictive maintenance algorithms.
Remote operations: In addition to supporting API calls for reading data from operational equipment, Synse can also be used to control equipment remotely. Vapor IO uses this capability of Synse to implement its remotely operable lights-out data centers.