Daniel Ninow, Solutions Consultant
It’s well known that within large IT datacenters, specialists in different domains (i.e. application, DB, networking, storage, etc) often struggle with sharing tools to solve problems. Vendors produce tools that may give a deep view of one domain, but lack correlation with others.
A national medical practice and research foundation customer, servicing over a million patients a year recently shared a real-world example of how VirtualWisdom provides insight to every layer of the infrastructure stack leading to rapid issue resolution. A database administrator reported an inexplicable problem that a storage administrator was able to diagnose, through a heuristic analysis and determine probable cause with recommended corrective actions for the VMware team to implement.
As is often the case, the origin of the issue came from someone reporting slowness from an app running on an SQL database. The VMware team saw evidence of a ‘CPU wait’ issue from the native VMware tools, but couldn’t nail down the real cause. So the storage infrastructure team fired up VirtualWisdom and, using the ProbeVM software integration with ESXi, they quickly created a report that covered ‘CPU utilization’, ‘CPU wait’, and ‘memory utilization’. The VMware tools were only keeping data for a week, so they used VirtualWisdom to go further back in time, to discover what had changed to cause the increased ‘CPU wait’. With the longer view, they correlated the cause to the addition of workloads and additional CPUs several days back. After pulling back the additional CPUs, they re-ran the reports and the ‘CPU wait’ had returned to acceptable levels. As you VMware admins know, you can negatively impact the performance of your environment since the VMkernel has to try and find a time slot in which it can give all vCPUs access to the physical cores.
Through VirtualWisdom’s intuitive AI-enabled analytics, the rogue application slowdown was quickly and easily identified, relayed to the appropriate triage team and resolved. In addition to significant CAPEX/OPEX infrastructure optimization savings, VirtualWisdom improves operational efficiency through proactive issue resolution and dramatically accelerated troubleshooting.