In Server & Application Monitor (SAM) 2019.4 we released hardware health monitoring for Nutanix clusters, alongside a technical preview of basic cluster monitoring. In the same version, we also released our first version of API polling as well with support for basic authentication! Nutanix has a rich ecosystem of APIs available for monitoring. How many of you tried to monitor your Nutanix environment?
When I tried with the 2019.4 API poller, I was immediately needing a way to skip the SSL certificate verification. This was also noted by customers and the feature requests started coming in.
Feature Request: Skip SSL Certificate Validation
I'm happy to say that in SAM 2020.2, this feature request has been fulfilled. And, even better, 2020.2 is Generally Available! The Nutanix monitoring provided is no longer in technical preview and with Virtualization Manager (VMAN) 2020.2 incredibly robust.
Regardless, APIs tend to be updated very frequently, so this walkthrough will help you supplement the out-of-the-box monitoring provided in SAM & VMAN without needing to update to later releases.
The Nutanix API documentation I'm referencing for these steps can be found here: https://www.nutanix.dev/reference/prism_element/v2/topics/authentication/
First, configure your Nutanix monitoring by adding a Nutanix cluster to be monitored, and configure hardware health polling too!
Navigate to your management widget on the AHV node of your choice.
Click 'Create'
While configuring the API poller, simply toggle that SSL certificate verification to disabled and you're now off to the races.
In the following screenshot I've highlighted a few changes to start playing around with as well now that you will have responses coming in. Quickly edit the title of your API poller, and monitor the status and response time of the API endpoint.
Let's start with url format:'any_cvm_ip:9440/.../clusters' to start taking a look at my Nutanix cluster.
Expanding out the response, I can immediately start monitoring the number of nodes for the cluster.
By setting a threshold, I can now be alerted if the number of nodes drops below 3, as this can impact my data resiliency status.
Hmm, I'd like to start drilling down into a few key metrics, like what are the read IOPs for the cluster. Here it looks like I have a few choices, I can continue within this same request and expand out the stats response, or I can create another request.
Let's take a quick look at another new addition that you can use to accelerate your API polling. Look in the upper right hand corner to see some new options
Clicking on the 3 dots will allow me to duplicate my request so I don't have to configure a new API poller and set up my authorization more than once.
Here's where I am, you can see I have 2 requests, identically configured.
Let's configure the second request with a different URL. 'any_cvm_ip:9440/.../stats'
Ta-da! Now I'm able to quickly monitor Read IOPs for the cluster.
Don't forget that with the API pollers, this is integrated with Orion Maps, PerfStack, and the new Orion Dashboards so you'll be able to not only quickly monitor additional data from your Nutanix environments, but visualize the results better than ever.
Across the Orion platform, you have multiple options to gain 100% monitoring insight for your Nutanix environment:
- Network Performance Monitor (NPM) allows visibility through the UnDP for access to metrics via SNMP
- Server & Application Monitor (SAM) gives you in depth hardware health, Application Monitor templates and the API poller for quick API access
- Virtualization Manager (VMAN) provides in depth and robust monitoring full stack support for Nutanix AHV, mixed VMware / Nutanix and Hyper-V / Nutanix environments
- Log Analyzer (LA) allows you to provide tagging and alerting on syslogs from your Nutanix environment and see those events side by side with other event types.
Not only that, we are Nutanix Ready Certified as a partner solution that can be run on the Nutanix AHV hypervisor and also utilizes the Nutanix REST APIs to work with AHV.
Let us know how your monitoring needs can be better served, and as always we look forward to your feature requests for future enhancements.