This morning when I logged in, I saw an alert from a test environment where we are constantly testing Applicare. I was informed of slowness in JPetStore transaction’s response time in the middle of the night.
I drilled down into those transactions and immediately saw that any backends/databases were not the cause of the issue. This helped me immediately eliminate them from the root cause analysis.
I could also see from Top Business Transactions that SQLs were getting processed fairly quickly and hence weren’t causing the delay.
By focusing on the JPetStore app server and host I could clearly see that # of processes running on the host has gone up during that period. I also saw disk transfers/sec has gone up significantly.
That made me look at what processes were running on the machine and which one of those processes was consuming resources. I saw MsMpEng – Windows Security component formerly called Windows Defender is consuming the resources and that caused the slowness in the transactions.
This is just an example scenario that we see in real life many times. We could have gotten to the same root cause results many other ways too e.g. from a higher CPU alert right in my inbox that includes top processes at the time of alert. But I choose the longer route to show the root cause analysis process which will come handy in many other scenarios.
Happy Applicare-ing for your apps and systems 🙂