We've analyzed thousands of your deployments and compared with Managed Hardware Requirements doc page. It seems that more IOPS are required for smooth running Web UI, REST API, on-demand analysis etc.
Solved! Go to Solution.
Managed hardware requirements have been updated for required IOPS per cluster node to reflect current needs to provide on-demand analysis. It's about 10x more.
Please adjust as required on your side. Also that's important to you if you plan new deployments to follow new guidelines.
For the reference I leave the previous version screenshot:
Thanks for your input! I've had a feeling for a long time that disk requirements were set too low. In my opinion, writing so much information to disk as it arrives causes a lot of disk activity, and it can lead to reading being slow. I'm eager to see if the rollout of 1.230 and it's monitoring tenant will contribute to a better knowledge of what is happening in out client Managed installations 🙂
It seems that I'll have to wait some more time to get those disk metrics, as the OneAgent info seems to not be available in the new self-monitoring environment. I have commented this at https://community.dynatrace.com/t5/Dynatrace-Open-Q-A/Local-self-monitoring-environment/m-p/177260
Will this be available in the near future?
This is awesome. We have been struggling with one of the customers about slowness in Dynatrace UI and we were troubleshooting this since last 6 months. Now this cleared my suspicion about what might causing the slowness. Trust me, this helped me in a big big time.