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
Currently we are making use of spanmetrics connectors for generating metrics. we observed high volume of metrics being generated with spanmetrics connector. It causes high datapoints ingestion to Victoria metrics backend. To test it we compared with tempo metrics generator. For testing we use same data pipelines on both parallelly, and we recorded performance with same query on spanmetrics connector as well as tempo metric generator. so, with tempo metric generator with "traces_spanmetrics_calls_total" metrics it produces 6706 results series whereas spanmetrics connector with "calls_total" metrics produces 13363 results series with same timestamp. Also, if we compare the data points ingestion to Victoria metrics backend so, in this case tempo metrics generator produces 7.18K mean whereas spanmetris connector produces 53.6K mean which is quite high. this datapoint ingestion is with 14.2K mean span rate received. so, in this case spanmetrics produces 16K mean metrics point rate. So, with this, is it we are missing any configuration? or is it some internal configuration needed?
Steps to Reproduce
Expected Result
The metrics generation by spanmetrics connector should match with tempo metrics generator or it should near to what tempo metrics generator producing with same setup.
Component(s)
connector/spanmetrics
What happened?
Description
Currently we are making use of
spanmetrics connectors
for generating metrics. we observed high volume of metrics being generated with spanmetrics connector. It causes high datapoints ingestion to Victoria metrics backend. To test it we compared with tempo metrics generator. For testing we use same data pipelines on both parallelly, and we recorded performance with same query on spanmetrics connector as well as tempo metric generator. so, with tempo metric generator with "traces_spanmetrics_calls_total" metrics it produces 6706 results series whereas spanmetrics connector with "calls_total" metrics produces 13363 results series with same timestamp. Also, if we compare the data points ingestion to Victoria metrics backend so, in this case tempo metrics generator produces 7.18K mean whereas spanmetris connector produces 53.6K mean which is quite high. this datapoint ingestion is with 14.2K mean span rate received. so, in this case spanmetrics produces 16K mean metrics point rate. So, with this, is it we are missing any configuration? or is it some internal configuration needed?Steps to Reproduce
Expected Result
The metrics generation by spanmetrics connector should match with tempo metrics generator or it should near to what tempo metrics generator producing with same setup.
Actual Result
Span rate recieved
Metric points rate
Datapoints ingestion rate to victoria metrics
Collector version
0.104.0
Environment information
No response
OpenTelemetry Collector configuration
Log output
No response
Additional context
No response
The text was updated successfully, but these errors were encountered: