So if I could image what monitoring would look like in the future I would want the solution to be “unicorns and rainbows” out-of-box. So what do I mean by this? I would want a solution that analyzes my environment for me and tells me exactly where my issues are based upon my environment. I would expect that this application monitoring solution would have to analyze my environment and its usage for some period of time to know what it should recommend. Another monitoring wish that I have is that I would be able to allow it to “fix” my server and host related issues for me. This level of application performance monitoring has the potential to create not only monitoring efficiency, but administrative efficiency too.
Server Virtualization in the Future
So can we expect our application monitoring tools to just magically fix all of our virtual server and host issues as we look to the future? I say, why not? We have discussed the ongoing challenges that come with deploying a monitoring solution correctly including application monitoring performance. If we can allow our software solution to prevent future problems by preparing and optimizing resource allocation with predictive analytics and self correction, then we should do this!
What types of solutions are we starting to see as it relates to the future state of monitoring?
The Future of Application Monitoring
- Auto adjustment of guest CPU would be extremely beneficial during cases of routine CPU contention. Additionally, it would also be important that the tool can choose a time to bring the guest down when no one is using it, or alternatively lets the administrator know about this recommendation allowing them to do the work at a suitable time.
- Similarly, automatic memory modification on virtual servers would be useful in cases where memory is needed to keep application performance in check. Potential challenges with this would be ensuring that paging file modifications would also be made, and system reboots would be coordinated at times that are appropriate for the organization.
- Advanced analytics into how my server resource utilization is spread across multiple hosts. So while I feel that there are some monitoring tools that provide some level of insight into this detail, there is still a growth opportunity into deeper analytics to support administrative level work.
- Leverage my monitoring tool to move guests to different hosts instead of built-in tools. Again there are some tools that do this including native tools that work really well. The opportunity here is that using one tool to do this ensures that the system isn’t trying to adjust CPU or memory at the same time a guest is moving to a new host. One tool to track all the changes and prevent conflict of the automation occurring.
- Receive recommendations on when to do guest maintenance to minimize/avoid downtime, but maintain optimal guest performance. As I imagine this as an administrator I would have insight through either a single pane of glass, or minimal clicks an understanding of how to optimize my overall virtual environment proactively.
What to Expect?
I would say that with the right amount of application monitoring and application monitoring performance commitment we could definitely expect all of these things of our monitoring tools. There are some products already getting started with these automation aspects today. The possibility that this could eliminate some of the complexity we experience today is really exciting to consider.
We really have come a long way with monitoring. More than a decade ago our options were minimalistic. Today while we face some of the same deployment challenges of years ago, the monitoring options when configured correctly will allow us to sleep peacefully at night. I do mean that literally as someone who spent almost 20 years in industry on-call. The more proactive the better.
Finally taking the time to image what is starting to take place with monitoring tools, and the automation possibilities for monitoring of the future.