Use Case
User can find cardinality contribution of label or combination of labels using cc mode. After that they have to decide the approach that they want to take, it may be dropping, aggregating or partitioning (relabelling). For taking any action, they have to do some sort of analysis like:
- If we drop the label is it going to result in duplicates?
- Are we using this label somewhere else may be in our grafana or alert manager?
Approach
Will add subcommand to cc mode to run the analysis as per action. Add 2 subcommands:
- drop
- This will give out of total timeseries after dropping the label how many of them are unique and give a % of that
- aggregate [type]
- provide type as Grafana/AM and then the token and url to find the queries where we are using the metric
Use Case
User can find cardinality contribution of label or combination of labels using cc mode. After that they have to decide the approach that they want to take, it may be dropping, aggregating or partitioning (relabelling). For taking any action, they have to do some sort of analysis like:
Approach
Will add subcommand to cc mode to run the analysis as per action. Add 2 subcommands: