Skip to content

Conversation

@ArthurSens
Copy link
Member

Description

This idea came up after talking with @bwplotka about the appenderv2 adoption. Is there a reason not to force ScrapeManager to convert classic histograms into NHCB? If we assume all histograms are ingested that way, we could eventually get rid of state handling for classic histograms.

I'm not adding a changelog entry because, theoretically, it has no impact on the end user.

Would love to get your opinions here @dashpole @krajorama

Signed-off-by: Arthur Silva Sens <arthursens2005@gmail.com>
@ArthurSens ArthurSens requested review from a team and dashpole as code owners January 23, 2026 12:46
@github-actions github-actions bot added the receiver/prometheus Prometheus receiver label Jan 23, 2026
@dashpole
Copy link
Contributor

I think this makes a ton of sense

@bwplotka
Copy link
Contributor

BTW: Future scrape protocols will only support this case better (e.g. OpenMetrics 2.0)

Copy link
Contributor

@bwplotka bwplotka left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Would be curious to grab some benchmarks to see if it's improving the efficiency here. I'd expect some benefits.

Signed-off-by: Arthur Silva Sens <arthursens2005@gmail.com>
@ArthurSens
Copy link
Member Author

ArthurSens commented Jan 23, 2026

I've got some weird test failures with this: start time and staleness markers disappearing from the converted NHCB. Not sure yet if this is expected, a bug in prometheus or if more code needs to be refactored here.

I couldn't find time to investigate yet though

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

receiver/prometheus Prometheus receiver

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants