JFrog Connect solves the major challenges of maintaining software in Linux-based IoT and edge devices, including fleet visibility, deploying software updates at scale, monitoring, and troubleshooting.
Fast, secure, and developer friendly, Connect automates software deployment from code to your Linux-based devices. The high-level introduction below describes how you can use Connect to solve these challenges and bring your code to the edge.
Connect brings IoT devices into the DevOps pipeline by covering the crucial segments of Deploy, Operate, and Monitor. Covering these segments for IoT means that you need to have full knowledge of your fleet all the time, have an easy deployment mechanism, and have full and secure control of your devices. Once you’ve covered these, scaling to hundreds of thousands of devices is easy.
Connect provides an easy-to-use web UI where you have visibility over your fleet at multiple levels of granularity.
At fleet level, you can see overview information including the online/offline status of devices, device locations, indications about live alerts, different versions in the field, and new device registrations. You can also filter the views to see this information per device name, group, or tags defined.
For any specific device, you can get a drill-down of its technical details including CPU, RAM, and disk space, and application information such as app versions running and logs available. These levels of visibility are key to scaling your edge and IoT projects.
You can use the Connect REST API to track information about your fleet, including all of the information available in the web UI. Using the REST API is particularly desirable when your fleet size is in the hundreds, thousands, or more.
Connect enables you to build robust and reusable update flows. You can build flows to suit any type of installation process by using the prefab action blocks that can run commands and scripts, deliver files and containers, and pull binaries and files from repositories such as Git and Artifactory. You can also parameterize data fields in your flows to enable dynamic deployments that will scale for your entire fleet. In case a failure occurs at some point in the update, Connect has a configurable mechanism to roll back per action block or roll back the entire update.
You can monitor CPU, RAM, and disk space of each device and monitor the Up/Down status of processes on the device. Using the alerts tool, you can configure the type of alert (resource or process), and define the thresholds and boolean logic that will trigger an alert. Using a simple webhook, alerts can be integrated with other applications such as email, Slack, and PagerDuty.
The Connect web UI makes it easy to troubleshoot and maintain devices by using remote support tools. With Connect’s control center, you can run commands remotely on a device or connect to it for full access and control. This enables you to troubleshoot the device securely from your own desktop. Connections with the device employ secure port forwarding (e.g., secure remote SSH, VNC, webview) for any interface connection, as well as reverse SSH.
Managing a large fleet of devices and different operating environments takes some organization. Connect makes it easy for you to segment your device fleet into any divisions and hierarchy that suits your business.
The highest organizational level is the Project, in which you can configure completely different environments with different device types and use cases. For example, you could configure different communication cycles per project. At this level, you can also configure which users have access to which devices.
Within a project, you can define groups and sub-groups of devices. This level is good for differentiating between administrative and geographic areas.
Tags offer another method of grouping, enabling you to create subsets of devices based on any characteristics you choose.
Once you have familiarized yourself with the Connect capabilities, you will be ready to take your deployments to the next level by automating and integrating Connect into your DevOps pipeline. You can use Connect’s REST API to fully automate your software deployments, and Connect is already fully integrated with key pipeline links such as Docker and Docker Compose for containerized apps, and the Git and JFrog Artifactory repositories. If you use Connect together with the full JFrog Software Supply Chain, you will get the full power of the JFrog platform.
The Connect user documentation provides: