docs: correct tcp_connection_remote label documentation - #2608
Open
xRookieFight (xRookieFight) wants to merge 1 commit into
Open
docs: correct tcp_connection_remote label documentation#2608xRookieFight (xRookieFight) wants to merge 1 commit into
xRookieFight (xRookieFight) wants to merge 1 commit into
Conversation
Signed-off-by: xRookieFight <me@xrookiefight.com>
xRookieFight (xRookieFight)
requested review from
Evan Baker (rbtr) and
Vipul Singh (vipul-21)
August 4, 2026 14:33
Author
|
@microsoft-github-policy-service agree |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Description
Fixes #2559.
The docs describe
tcp_connection_remoteas a per-remote-endpoint metric withaddress(IP:port) labels, but that hasn't been true since #953, which deliberately dropped the port label and folded every remote socket into a singleaddress="AllIPs"series to keep the metric's cardinality bounded.So this is a docs bug rather than a code bug - the aggregation is the intended behavior. This PR takes option 2 from the issue and makes the documentation match what the exporter actually emits:
docs/03-Metrics/02-hubble_metrics.mdanddocs/03-Metrics/modes/basic.md: label column now saysaddressis alwaysAllIPs, plus a short note in the basic-mode label values section explaining why individual remotes aren't exported.docs/06-Troubleshooting/basic-metrics.md: the sample scrape output still showed per-IP series with aportlabel that no longer exists. Replaced with the single aggregated series.pkg/metrics/types.go: the gauge's HELP string said "by remote address", which is the same misleading claim shipped to every Prometheus scrape. Reworded.pkg/plugin/linuxutil/netstat_stats_linux.go: added a note on theaddrDefaultTCPRemoteconstant so the aggregation doesn't read like an unfinished implementation next time.Worth calling out for the issue author: this closes the docs mismatch but does not give you the per-remote breakdown you were after. If there's appetite for that as an opt-in (default stays aggregated, a config flag turns on per-endpoint series with stale-series cleanup between polls), I'm happy to follow up with a separate PR - but that seemed like a decision for the maintainers rather than something to sneak into a docs fix.
Related Issue
#2559
Checklist
git commit -S -s ...).Screenshots (if applicable) or Testing Completed
No behavior change, so nothing new to test -
go build ./pkg/metrics/... ./pkg/plugin/linuxutil/...passes and the only runtime-visible difference is the HELP text on the gauge.