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
To collect metrics about the Elasticsearch indices Metricbeat calls the /_stats endpoint. Inside the object "indices" : { all indices are listed with its names. The data that is missing inside is the index uuid. It can be fetched by calling GET /index-name and then can be found under uuid. Having to make a separate request for each index to fetch the identifier is not an optimal solution.
The uuid is important for the following reasons:
The same index name can exist in multiple clusters. Beats can collect data from multiple clusters into a single monitoring cluster.
An index can be deleted and a new one created with the same name. In this case only the UUID is different.
With cross cluster replication matching indices from 2 different clusters with the same uuid could become even more interesting: Introduce cross-cluster replication #30086
NOTE: The _stats endpoint also currently does not expose the cluster_uuid and we fetch it in a separate call.
To collect metrics about the Elasticsearch indices Metricbeat calls the
/_statsendpoint. Inside the object"indices" : {all indices are listed with its names. The data that is missing inside is the index uuid. It can be fetched by callingGET /index-nameand then can be found underuuid. Having to make a separate request for each index to fetch the identifier is not an optimal solution.The uuid is important for the following reasons:
NOTE: The _stats endpoint also currently does not expose the
cluster_uuidand we fetch it in a separate call.