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
The plugin works fine, but unfortunately the way it structures the events that it creates is less than ideal.
In my particular case, I tried to use the Netapp Harvester (which gets metrics from a Netapp storage system and outputs plaintext Graphite) with this input plugin so as to not have to spin up an actual Graphite server.
The result was a bunch of events with field names like: netapp.capacity.Cluster-HQ.Cluster-HQ.node.Cluster-HQ-01.aggr.cluster_hq_01_aggr1.snapshot_reserve_avail
Elasticsearch quickly began to throw errors when the number of fields in the (brand new, empty) index hit 1000. It occurs to me that any reasonable Graphite deployment using a metrics collector would similarly hit those kinds of numbers.
What I would like to do is to have some options to re-map the name.
Option 1 could be as simple as just having a field named 'metric' containing the metric name, and a separate field called value with the value. May be some issue with conflicts for non-numeric values, might have to go with a text or similarly named field for those instead of value. If nothing else, this would make the plugin perfectly useful for me.
Option 2 would be more complex, you'd split on dots, and then with the resulting data create a hierarchy, so you'd map from:
The plugin works fine, but unfortunately the way it structures the events that it creates is less than ideal.
In my particular case, I tried to use the Netapp Harvester (which gets metrics from a Netapp storage system and outputs plaintext Graphite) with this input plugin so as to not have to spin up an actual Graphite server.
The result was a bunch of events with field names like:
netapp.capacity.Cluster-HQ.Cluster-HQ.node.Cluster-HQ-01.aggr.cluster_hq_01_aggr1.snapshot_reserve_avail
Elasticsearch quickly began to throw errors when the number of fields in the (brand new, empty) index hit 1000. It occurs to me that any reasonable Graphite deployment using a metrics collector would similarly hit those kinds of numbers.
What I would like to do is to have some options to re-map the name.
Option 1 could be as simple as just having a field named 'metric' containing the metric name, and a separate field called
value
with the value. May be some issue with conflicts for non-numeric values, might have to go with atext
or similarly named field for those instead ofvalue
. If nothing else, this would make the plugin perfectly useful for me.Option 2 would be more complex, you'd split on dots, and then with the resulting data create a hierarchy, so you'd map from:
netapp.capacity.Cluster-HQ.Cluster-HQ.node.Cluster-HQ-01.aggr.cluster_hq_01_aggr1.snapshot_reserve_avail = 100
to
This option would enable the canny Elasticsearcher to effectively be able to query the hierarchy the same way as a Graphiter would with wildcards.
The text was updated successfully, but these errors were encountered: