>> Hi, I'm Michael, and I'm a Product Manager on
the IoT Edge Team. Allow me to demonstrate just a
few of the things that are now possible to integration that
we've done to Azure Monitor. This integration is
based on a set of built in metrics that we
provide out of the box, a new collector module and a set of curated visualizations that
are available by default. Now, to give you an example of how Azure Monitor allows you
to manage your fleet, I'd like to show you how you can use alerts to respond to anomalies. A new application was recently deployed to one of
my IoT Edge devices. It filters the messages from a temperature sensor before they're
forwarded upstream to IoT Hub. I already had in place a number of alerts to tell me when
something goes wrong, and one of those alerts checks for a high queue length in
the Edge Hub component. The check for this alert watches to see if any of the message
queues at the edge of maintains have grown beyond what I've considered
unacceptable threshold. Now, earlier today, I
received this e-mail alert indicating that one of my devices encountered this High
Queue Length condition. Let's take a look at
the affected resource and see what's going on. Under the Monitoring section, I'll find a set of default
Workbooks for my IoT Edge devices. These presents some curated
visualizations that I can use. Let's start with the Alerts View. In this view, I can see all of the alerts that are active
across all of my IoT Hubs. Right now, there's just the one. I can quickly see that
on this Edge01 device, I've hit this High
Queue Length condition, and clicking on the
device name is going to take me to a details view. In this view, I immediately see that the problem is with
my Filter Function. It has the queue length
that is growing. If I select it, I can
even see that over time that's what appears
to be happening. If we look at the sender, we can also get a sense of
how the messages are flowing from the sender to
the run-time in blue. From the run-time to
the parent receiver, the filter function in orange, where it doesn't seem to be receiving and processing
those messages. At this point, let's troubleshoot
and look at the logs. In this view, we can
pull the logs directly from one of the modules running on the device to see
what's happening. Now, let's check the
filter function. Right away it looks like there's a Stack Trace of an exception is being thrown in the
code for that model. That most likely explains
what's going on and why the messages aren't
being processed. At this point, I really
need to go talk to the Developer and figure
out what's going on. While this has been
useful for drilling in and troubleshooting a single device, what if I want to get a
sense of the health for all of my devices on my IoT Hub? For that, we have the Fleet View. In this Fleet View
we get a summary for the past two hours by default of
the devices that are in my fleet. Now this is just the
current IoT hub, but if I have other IoT hubs, I can include those as well. I can drill into the
list of devices, get a snapshot for the health of
a given device and if I need to, I can jump back to that detailed
view that we saw earlier. In closing, what I've shown
is available out of the box, but it can also be extended or customized to suit
your particular needs. We believe that this
integration with Azure Monitor will help
you get more out of your IoT Edge devices and
make troubleshooting and monitoring that much easier. Thanks.
