Index cluster.id and cluster.name in elasticsearch/ml_job metricset #9165
Conversation
ycombinator
added
review
Metricbeat
|
|
||
| // +build !integration | ||
|
|
||
| package ml_job |
There was a problem hiding this comment.
don't use an underscore in package name
There was a problem hiding this comment.
As mentioned in other PR's, worth adding a changelog I think.
General note: Right now the code is in each metricset which is ok. To simplify the code perhaps have something like
func enrichEvent(info elasticsearch.Info, event beat.Event) beat.Event { ... }
Like this it would be only 1 line in each metricset and if we decide to add more general fields in the future we can do it in one place. That is basically what I meant in the past with a "processor".
I think in general yes but these fields (cluster_uuid and cluster_name) were a miss. They should've been there to begin with because some other metricsets in this module have them but some don't. I say we add CHANGELOG entries for new fields once the module goes GA?
Ah yes, I totally forgot about this idea 🤦♂️. Will refactor all of these in a follow up PR! |
ycombinator
deleted the
metricbeat-elasticsearch-ml-job-cluster-id-name
branch
ycombinator
removed
the
needs_backport
…csearch/ml_job metricset (#9173) * Index cluster.id and cluster.name in elasticsearch/ml_job metricset (#9165) This PR teaches the `elasticsearch/ml_job` metricset to index the Elasticsearch `cluster_uuid` and `cluster_name` as the module-level `cluster.id` and `cluster.name` fields, respectively. (cherry picked from commit e7f4a1c) * Fixing field name for 6.x
This PR teaches the
elasticsearch/ml_jobmetricset to index the Elasticsearchcluster_uuidandcluster_nameas the module-levelcluster.idandcluster.namefields, respectively.