Part 3 in a 4-part series
By Ravi Prakash, Director Product Marketing
Value of an agentless approach to compute infrastructure
In a virtualized compute environment, you have the issue of noisy neighbors – a VM from a less mission critical tier vying for resources with a VM associated with a more critical tier 0 application. While VMware provides tools like Distributed Resource Scheduler (DRS) which wait for a problem to occur before moving a VM from one ESX host to another using VMware VMotion, what if you could proactively prevent the resource contention problem in the first place?
This is addressed by our custom analytic VM Coordinator. It uses artificial intelligence (AI) and machine learning to learn VM workloads and then recommend optimal placement of VMs in an ESX cluster to proactively avoid memory, disk or CPU contention. It uses the history of your workloads to understand peaks and valleys, so you may never have to use DRS.
Value of mapping applications to infrastructure
As a user of VMware, if you are using vSphere Distributed Switch(vDS), VirtualWisdom can be configured as a NetFlow collector for NetFlow generated by vDS. By observing inter-VM communications VirtualWisdom can deduce which VMs contain components of the same application.
VirtualWisdom has progressed beyond infrastructure monitoring into the area of application detection using NetFlow.
If you are using Hyper-V or PowerVM instead of vSphere, VirtualWisdom can detect applications by interacting with the Change Management Database (CMDB) of ServiceNow, if you are using ServiceNow.
The benefits of doing this include:
Value of adding “infrastructure awareness” to Application Performance Management (APM)
If you have AppDynamics or Dynatrace to monitor tier 0 applications why do you need VirtualWisdom? For the simple reason that finite budgets typically restrict you to instrumenting APM tools for only 10-15% of your applications.
How do you deal with the fact that any of the remaining 85% of tier 2 or tier 3 applications sharing the same underlying infrastructure can impact the SLA of your tier 0 application? VirtualWisdom helps you here.
Need context? Read Part 1! and stay tuned for part 4 of this 4-part series where we’ll cover root-cause using cross domain correlation and run-book automation.