The Cloud proxy logs ship to Elasticsearch logging clusters including the x-elastic-client-meta header for all Elasticsearch requests. In order to capture and aggregate CLI telemetry data, this configuration needs to be extended to work for ALL HTTP requests, not just Elasticsearch.
This will require working with the Platform Observability team to get changes reviewed, merged and shipped. Due to the sheer volume of HTTP proxy logs Cloud indexes every day, there are limitations to how much new data they are comfortable adding to the indexed logs. There is also a possibility that proxy infrastructure is different for Elasticsearch vs. other APIs like Kibana or the control plane, so some discovery is still needed on that point.
Open questions:
The Cloud proxy logs ship to Elasticsearch logging clusters including the
x-elastic-client-metaheader for all Elasticsearch requests. In order to capture and aggregate CLI telemetry data, this configuration needs to be extended to work for ALL HTTP requests, not just Elasticsearch.This will require working with the Platform Observability team to get changes reviewed, merged and shipped. Due to the sheer volume of HTTP proxy logs Cloud indexes every day, there are limitations to how much new data they are comfortable adding to the indexed logs. There is also a possibility that proxy infrastructure is different for Elasticsearch vs. other APIs like Kibana or the control plane, so some discovery is still needed on that point.
Open questions:
x-elastic-client-metaheaders in regional logging clusters for non-Elasticsearch traffic (e.g. cloud control plane + Kibana)?