In this second installment, I’ll show specifically how Foglight can help monitor the Dell Edge Gateways that are at the core of our Internet of Things reference architecture.
In case you missed it, part 1 is available here.
There are many components that Foglight can monitor in an IoT environment. These include monitoring the performance, availability and latencies of the Dell Edge Gateways (or other hosts), the networks linking them back to your datacenter, and the URL’s that are used to send data to the cloud for analysis. Of course, Foglight can also monitor your other applications, databases and infrastructure.
Foglight’s modular architecture opens up monitoring to an IoT environment.
The Dell Edge Gateway 5000/5100 runs Linux or Windows 10 and is “just another host” that can be monitored by Foglight.
It can take a deep-dive into the gateway itself, specifically on Linux, monitoring cpu/memory/network/disk along with top processes, disks, etc. The same can be done for a SonicWall or Force10 switch, where it can report back snmp-based data (interfaces, utilization, etc.)
The Dell Edge Gateway shows up in Foglight like any other Linux host would.
Adding in snmp monitoring of the Dell Edge Gateway allows you to monitor network interfaces and running processes.
Both the infrastructure and snmp monitors can poll the Dell Edge Gateway from a remote host so that nothing needs to be installed on the device itself. The gateway has snmp turned on by default (Linux) and is available with Windows 10 IoT Edition as well.
In the next post, we’ll look at monitoring the network and web links that the Dell Edge Gateway uses to connect devices and send data to the cloud. Until then, here is another look at what we’re working towards – a custom IoT dashboard.
Here is the end result – a dashboard showing important metrics and data from our Dell Edge Gateway that was used for customer trials in our Santa Clara IoT lab. This was done using the “drag ‘n drop” dashboarding interface that is available to all Foglight users. Finding and sizing an appropriate background image was the longest task – after that was uploaded, it took approximately 10 minutes to find the desired data and refine how it is presented.