-
Notifications
You must be signed in to change notification settings - Fork 2.4k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[processor/transform] Metric conversion functions use datapoint context instead of metric context #20773
Comments
Pinging code owners:
See Adding Labels via Comments if you do not have permissions to add labels yourself. |
At the time of their creation, which was during transformprocessor/OTTL infancy, we didn't have the concept of the metrics context, we only had spans, datapoints, and log records. You are correct that these are better suited for the metrics context. |
That makes sense. Are we OK with changing these to be on the metrics context (with a feature flag, since it would be a breaking change)? I'd be willing to take that on, if you think so. |
As long as we follow the feature gate deprecation process go for it. |
@BinaryFissionGames I would suggest waiting until #18822 is merged to start your work. |
Got it, thanks for the heads up on that. |
This issue has been inactive for 60 days. It will be closed in 60 days if there is no activity. To ping code owners by adding a component label, see Adding Labels via Comments, or if you are unsure of which component this issue relates to, please ping Pinging code owners:
See Adding Labels via Comments if you do not have permissions to add labels yourself. |
This issue has been inactive for 60 days. It will be closed in 60 days if there is no activity. To ping code owners by adding a component label, see Adding Labels via Comments, or if you are unsure of which component this issue relates to, please ping Pinging code owners:
See Adding Labels via Comments if you do not have permissions to add labels yourself. |
@BinaryFissionGames are you still interested in working on this? It should be ready for implementation. |
Sorry for the late response @evan-bradley - I don't think I'll be getting around to this anytime soon, so if anyone else has any interest in taking it, they should totally go for it. |
…#29091) **Description:** Allow running OTTL `convert_sum_to_gauge` and `convert_gauge_to_sum` in metric context instead of datapoint when `processor.transform.ConvertBetweenSumAndGaugeMetricContext` is enabled closes #20773 <!-- **Testing:** <Describe what testing was performed and which tests were added.> **Documentation:** <Describe the documentation added.> --> This is the result of an effort at contribfest at Kubecon NA '23. --------- Signed-off-by: Alex Boten <aboten@lightstep.com> Co-authored-by: Faith Chikwekwe <faithchikwekwe01@gmail.com> Co-authored-by: Tyler Helmuth <12352919+TylerHelmuth@users.noreply.github.com> Co-authored-by: Andreas Thaler <andreas.thaler01@sap.com> Co-authored-by: Antoine Toulme <antoine@toulme.name> Co-authored-by: Etienne Pelletier <etienne.pelletier@segment.com> Co-authored-by: bryan-aguilar <46550959+bryan-aguilar@users.noreply.github.com> Co-authored-by: Rajkumar Rangaraj <rajrang@microsoft.com> Co-authored-by: Curtis Robert <crobert@splunk.com> Co-authored-by: Alex Boten <aboten@lightstep.com> Co-authored-by: Jacob Marble <jacobmarble@gmail.com> Co-authored-by: Jon <jonnywamsley@gmail.com> Co-authored-by: Daniel Jaglowski <jaglows3@gmail.com> Co-authored-by: Antoine Toulme <antoine@lunar-ocean.com>
…open-telemetry#29091) **Description:** Allow running OTTL `convert_sum_to_gauge` and `convert_gauge_to_sum` in metric context instead of datapoint when `processor.transform.ConvertBetweenSumAndGaugeMetricContext` is enabled closes open-telemetry#20773 <!-- **Testing:** <Describe what testing was performed and which tests were added.> **Documentation:** <Describe the documentation added.> --> This is the result of an effort at contribfest at Kubecon NA '23. --------- Signed-off-by: Alex Boten <aboten@lightstep.com> Co-authored-by: Faith Chikwekwe <faithchikwekwe01@gmail.com> Co-authored-by: Tyler Helmuth <12352919+TylerHelmuth@users.noreply.github.com> Co-authored-by: Andreas Thaler <andreas.thaler01@sap.com> Co-authored-by: Antoine Toulme <antoine@toulme.name> Co-authored-by: Etienne Pelletier <etienne.pelletier@segment.com> Co-authored-by: bryan-aguilar <46550959+bryan-aguilar@users.noreply.github.com> Co-authored-by: Rajkumar Rangaraj <rajrang@microsoft.com> Co-authored-by: Curtis Robert <crobert@splunk.com> Co-authored-by: Alex Boten <aboten@lightstep.com> Co-authored-by: Jacob Marble <jacobmarble@gmail.com> Co-authored-by: Jon <jonnywamsley@gmail.com> Co-authored-by: Daniel Jaglowski <jaglows3@gmail.com> Co-authored-by: Antoine Toulme <antoine@lunar-ocean.com>
Component(s)
processor/transform
Describe the issue you're reporting
The function
convert_gauge_to_sum
,convert_sum_to_gauge
,convert_summary_sum_val_to_sum
andconvert_summary_count_val_to_sum
all are using thedatapoint
context, but I'm not sure I understand why.They all operate on the top-level "metric", so I think they should be using the
metric
context, since they should each run only once for each metric, and not once for each datapoint.It seems like this was just retaining the old behavior before contexts were introduced, but I think these functions are actually better suited on the
metric
context.The text was updated successfully, but these errors were encountered: