What is the best index_mode for the .Net trace data?

Elastic Observability APM
1

We collect trace data from a dotnet application with opentelemtry. Everything looks fine as expected. The use cause is checking the Observability => APM => Service whenever there is an issue or performance degradation.

My question is related to index_mode settings. The current and default index_mode is Standard for traces-apm@template. I wonder the followings.

  • Is it makes sense to switch to LogsDB or time series index mode?
  • What is the important points that I should check before change the index mode?

Cluster information:

  • The CPU usage on hot nodes is ~%20 on average.
  • There are two hot nodes and a frozen node.
  • ILM is set to keep data 7 days in hot nodes and move to the frozen node and remove data after 90 days.

Install method: Elastic Cloud
Elasticsearch version: 9.0.0
APM Server version: 9.0.0
APM Agent language and version: Opentelemetry | OTel SDK version -1.11.1
Elasticsearch subscription level: Enterprise

image
image1920×948 100 KB

2

Hi @Musab_Dogan good question
traces-* should be left in standard mode... they have not been test / not support with logsdb mode.

2 Likes