You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hello! Thanks for opening this ticket, can you open a support ticket here? This will give us a lot more visibility into other metrics, grab agent flares etc so we can narrow down exactly the cause of the memory increase here.
Ahh sorry for directing you to the support endpoint so quickly but I believe I've found a likely culprit:
Enables a cache for SQL and MongoDB obfuscation. This cache is enabled by default but can be disabled by setting apm_config.obfuscation.cache.enabled to false.
This feature can significantly improve the throughput of the trace-agent by reducing the CPU cost of obfuscating SQL and MongoDB queries however it does add a bit of additional memory overhead. I still recommend opening a support ticket so we can confirm for sure if this is the issue.
If you don't want the memory overhead / you aren't seeing CPU improvements or throughput improvements (which is possible depending on your exact workload) you can disable the cache using apm_config.obfuscation.cache.enabled.
Thank you @ajgajg1134, that sounds like a reasonable explanation and given the information I suspect this would have been an expected outcome so not really a bug. I will check to see if the respective setting will affect the memory utilization.
As you can see, there doesn't appear to be any visible positive impact on CPU utilization (average), thus it might make sense for us to disable the setting.
Agent Environment
Describe what happened:
Observing higher average memory utilization for the
trace-agent
after upgrading to v7.61.0.Describe what you expected:
Average memory utilization should not have increased after the upgrade.
Steps to reproduce the issue:
Observe the memory utilization metrics before and after upgrading to 7.61.0
Additional environment details (Operating System, Cloud provider, etc):
Datadog Agent running on AWS EKS, Kubernetes v1.31.1, Cilium v1.16.4.
The text was updated successfully, but these errors were encountered: