Add some basic metric to track metric latency and count #37
Conversation
| @@ -19,3 +19,8 @@ vector-common = { path = "../vector-common", features = ["btreemap"] } | |||
|
|
|||
| [build-dependencies] | |||
| prost-build = "0.12" | |||
|
|
|||
| [dependencies.tracing] | |||
There was a problem hiding this comment.
Is this also used in other modules?
There was a problem hiding this comment.
yes, this tracing is used in multiple source, such as :
Line 32 in 60673c7
| @@ -78,6 +84,10 @@ fn reparse_groups( | |||
| for (key, metric) in metrics { | |||
| let tags = combine_tags(key.labels, tag_overrides.clone()); | |||
|
|
|||
| let shard_name = tags.get("shardName").unwrap_or("unknown").to_string(); | |||
There was a problem hiding this comment.
Is it possible to make this shardName configurable at the source level?
There was a problem hiding this comment.
Wondering you mean configurable in source level is from this metric configuration? This metric should already have "xxx--general_1" shardName from scraping. This is a label for input source shardName we want to understand the load volume from each input shard. such nephos_1 is sending x series/seconds.
There was a problem hiding this comment.
I mean, having a new option in https://vector.dev/docs/reference/configuration/sources/prometheus_scrape/ since shardName is very databricks specific. It could be called latency_label and set to "shardName"
There was a problem hiding this comment.
You also need to update https://github.com/databricks/vector/blob/v0.39/README.databricks.md
And note the code should be merged into v0.39-custom
Improve visibility in vector metric path
count type:
latency: