LOG: Logs context filter selection logic #9147
Description
Improvement over #8504 (comment).
One we released this, we have seen then some users have k8s, cloud, host in resource attributes even this all selected results in a query which exceeds the limit.
We should come with a strategy.
- Always include -
service.nameanddeployment.environment/env/environmentif present in resource filters. - Select only one from the list, if k8s is found then don't move to the next. And from k8s just pick the first found
{
"kubernetes": {
"priority_order": [
"k8s.pod.uid",
"k8s.pod.name",
"k8s.deployment.name"
],
"selection_rule": "pick_first_available"
},
"cloud": {
"priority_order": [
"cloud.resource_id",
"cloud.provider",
"cloud.region"
],
"selection_rule": "pick_first_available"
},
"host": {
"priority_order": [
"host.id",
"host.name"
],
"selection_rule": "pick_first_available"
},
"container": {
"priority_order": [
"container.id",
"container.name"
],
"selection_rule": "pick_first_available"
}
}
If the resource filter is still empty after the above. We keep this logic #8504 (comment) and remove the k8s, cloud, host, container from the logic as it is already added above.
Metadata
Metadata
Assignees
Labels
No labels