Skip to content
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

Add metric tag family for too many requests http errors #402

Open
wants to merge 4 commits into
base: develop
Choose a base branch
from
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 5 additions & 0 deletions changelog/@unreleased/pr-402.v2.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,5 @@
type: improvement
improvement:
description: Add metric tag family for too many requests http errors
links:
- https://github.com/palantir/conjure-go-runtime/pull/402
19 changes: 12 additions & 7 deletions conjure-go-client/httpclient/metrics.go
Original file line number Diff line number Diff line change
Expand Up @@ -50,13 +50,14 @@ var (
MetricTagConnectionNew = metrics.MustNewTag("reused", "false")
MetricTagConnectionReused = metrics.MustNewTag("reused", "true")

metricTagFamily1xx = metrics.MustNewTag(metricTagFamily, "1xx")
metricTagFamily2xx = metrics.MustNewTag(metricTagFamily, "2xx")
metricTagFamily3xx = metrics.MustNewTag(metricTagFamily, "3xx")
metricTagFamily4xx = metrics.MustNewTag(metricTagFamily, "4xx")
metricTagFamily5xx = metrics.MustNewTag(metricTagFamily, "5xx")
metricTagFamilyOther = metrics.MustNewTag(metricTagFamily, "other")
metricTagFamilyTimeout = metrics.MustNewTag(metricTagFamily, "timeout")
metricTagFamily1xx = metrics.MustNewTag(metricTagFamily, "1xx")
metricTagFamily2xx = metrics.MustNewTag(metricTagFamily, "2xx")
metricTagFamily3xx = metrics.MustNewTag(metricTagFamily, "3xx")
metricTagFamily4xx = metrics.MustNewTag(metricTagFamily, "4xx")
metricTagFamily5xx = metrics.MustNewTag(metricTagFamily, "5xx")
metricTagFamilyOther = metrics.MustNewTag(metricTagFamily, "other")
metricTagFamilyTimeout = metrics.MustNewTag(metricTagFamily, "timeout")
metricTagFamilyTooManyRequests = metrics.MustNewTag(metricTagFamily, "too-many-requests")
)

// A TagsProvider returns metrics tags based on an http round trip.
Expand Down Expand Up @@ -154,6 +155,10 @@ func tagStatusFamily(_ *http.Request, resp *http.Response, respErr error) metric
return metrics.Tags{metricTagFamily2xx}
case resp.StatusCode < 400:
return metrics.Tags{metricTagFamily3xx}
case resp.StatusCode == 408:
return metrics.Tags{metricTagFamilyTimeout}
case resp.StatusCode == 429:
Copy link
Contributor

Choose a reason for hiding this comment

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

This change makes sense, but I am a little worried about changing the semantic meaning of 4xx to except 408 and 429. Do we think this could be confusing?

Choose a reason for hiding this comment

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

We already treat client-side timeouts differently (see https://github.com/palantir/conjure-go-runtime/blob/develop/conjure-go-client/httpclient/metrics.go#L148), but these are not tied to any specific response status code, so not that applicable. I think the value of distinguishing 408s and 429s across other 4xx status codes is significant enough to justify this change. One option to keep both would be to keep 4xx tag, and add Timeout and TooManyRequests as additional tags, this way we don't break anything.

return metrics.Tags{metricTagFamilyTooManyRequests}
case resp.StatusCode < 500:
return metrics.Tags{metricTagFamily4xx}
case resp.StatusCode < 600:
Expand Down