The point of building and running your SDDC is to run applications, so it only makes sense to have some visibility into those applications. Without awareness of the health and availability of applications it is very difficult to measure and meet SLAs for your business or organization. It can also be difficult to identify and resolve issues that impact applications, lengthening the time it takes to repair problems. Many times, application monitoring is done using different tools than infrastructure monitoring, if at all.
Unfortunately, this means you have frustrated internal and external customers, and loss of credibility for IT Operations.
Before I came to VMware, I oversaw virtual infrastructure and storage for a large financial institution. I’ve been in the eye of the storm. I know what it is like to be blamed for outages or performance problems that aren’t necessarily my fault. “My application is slow!” is a phrase that still gives me chills because the implication is “and IT is to blame!”
In this blog post, I will show you how vRealize Operations 7.5 with Application Monitoring provides application awareness in your SDDC operations for end-to-end visibility. This will reduce the mean time to repair outages, give you the ability to better support application SLAs and increase the value of IT Operations.
Agent Lifecycle Management
Application Monitoring delivers full agent lifecycle management, including bootstrap deployment of Telegraf agents on virtual machines. It is easier than ever to monitor applications in vRealize Operations 7.5.
Notice I mentioned Telegraf, which is the collection agent used by Wavefront as well. This means that if you prefer to monitor applications using Wavefront, you can still benefit from Application Monitoring’s ability to deploy and manage Telegraf for Wavefront collection. So, whether you wish to vRealize Operations or Wavefront for application monitoring, you still get the benefit of centralized Telegraf agent management.
This allows for centralized control of agents, and agent configuration in the vRealize Operations Manager user interface using the new “Manage Agents” tab in the Administration > Inventory page.
Here you can install, uninstall, stop and start the agents. Supported services are automatically discovered and displayed, giving you the option to configure the agent to monitor those services.
Below is an example of configuring MySQL for monitoring. You see there are required and optional settings including the user defined display name, and the ability to activate and deactivate monitoring.
Most service configurations are similar, and requirements for monitoring are documented.
Full Stack Visibility
The point of all this is to give IT Operations end-to-end visibility from the OS and application services and components to the virtual machine and the supporting SDDC infrastructure. This is powerful and provides some key capabilities:
- Quickly determine if virtual infrastructure problems are impacting or could impact applications, and specifically which applications.
- Investigate application performance and availability issues in context to isolate the problem domain.
- Proactively assess if SDDC changes will impact applications or determine if recent changes to virtual infrastructure had positive, negative or neutral impact on hosted applications.
Additionally, remote monitoring extends the spectrum of troubleshooting visibility to external systems on which the application may have a dependency. For example, a third-party survey conducted at checkout could be monitored as part of the stack using an HTTP Remote Check to confirm availability, response time and expected result content. The screenshot below shows how a Remote Check
Going back to the earlier problem, “My application is slow!” can you see how having this information to provide not only proof of innocence for your SDDC but also help your application team understand the problem is related to a third-party provider would make you a hero?
Application Aware Troubleshooting
You may already be familiar with the popular troubleshooting dashboards included with vRealize Operations for virtual machines, hosts and other SDDC components. With 7.5 a new dashboard works with Application Monitoring to provide a workflow for troubleshooting starting from the application.
If you get the “My application is slow!” call, simply select the application from the list in the dashboard and the Advanced Relationship widget will display the application’s components, the guest OS and the supporting SDDC components. To the right of the relationships you will see KPIs for the selected application to make it easier for you to have a conversation with the application owner.
Other features of the dashboard include in context alerts and metric charts. With this dashboard, you can quickly determine the next logical path for troubleshooting by identifying bottlenecks to performance, failures leading to availability and eliminating SDDC components that are healthy.
Using Endpoint Operations Today?
No problem, you’re still covered in vRealize Operations 7.5 as Endpoint Operations is fully supported. However, as you may have guessed, Application Monitoring using the Telegraf agent is the future. If you are using Endpoint Operations, I recommend trying out Application Monitoring and moving over to it where possible.
While both Endpoint Operations and Application Monitoring are supported in vRealize Operations 7.5, running both concurrently on a VM guest OS is not supported. You should fully uninstall the Endpoint Operations agent before deploying the Telegraf agent via the ARC. Steps to remove the agent can be found in the public documentation.
Supported operating systems for Application Monitoring and Endpoint Operations in 7.5 differs slightly, so be sure to reference these links for more information. Also, application coverage in Application Monitoring includes 17 services, but we will be adding more to reach parity with Endpoint Operations.
If you use Remote Monitors in Endpoint Operations, those are also available in Application Monitoring as I described above.
To summarize, here are the differences in coverage between Endpoint Operations and Application Monitoring as of vRealize Operations 7.5:
Getting Started
If you like what Application Monitoring brings to vRealize Operations 7.5 and want to try it out, I’ve created a video which will step you through the process of deploying and configuring the Application Monitoring. If you are new to vRealize Operations, or you would prefer to test on a non-production environment, please download and install a 60-day trial.
As with Endpoint Operations, the Application Monitoring feature is available for OS monitoring with vRealize Operations Advanced while vRealize Operations Enterprise is required for full application support.
