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

Investigate removal of X-Amzn-Trace-Id for queuetime #631

Closed
jrothrock opened this issue Feb 25, 2021 · 2 comments · Fixed by #648
Closed

Investigate removal of X-Amzn-Trace-Id for queuetime #631

jrothrock opened this issue Feb 25, 2021 · 2 comments · Fixed by #648

Comments

@jrothrock
Copy link
Contributor

jrothrock commented Feb 25, 2021

https://docs.aws.amazon.com/elasticloadbalancing/latest/application/load-balancer-request-tracing.html

The X-Amzn-Trace-Id header that Amazon adds to ELB only gives time since epoch in seconds as opposed to milliseconds. This isn't a very accurate measurement, and seems to impact apps negatively (reporting times incorrectly, which seems to be a lot more pronounced since their recent change to hex decimal timestamps)

Whether Amazon is truncating the time or rounding the time to the second may need to be investigated, but either way unless queuetime is above a second this measurement doesn't add much value

@tim-schilling
Copy link
Collaborator

Good morning Jack. It looks like the header for the request is only going to have seconds. I could imagine them changing it to be a hexadecimal value to do milliseconds in the future, but I can't find any roadmap that would indicate as such. I agree that the appropriate measure is to remove this.

@tim-schilling
Copy link
Collaborator

Here's a amazon forum post requesting the feature that we'd need to restore this: https://forums.aws.amazon.com/thread.jspa?threadID=333252

tim-schilling added a commit to tim-schilling/scout_apm_python that referenced this issue May 19, 2021
The header's time value is in seconds which isn't enough precision
for us to provide quality data to users. With seconds precision, the
queue time hovers about 0.5s due to the truncation.

It may be possible to pull the millisecond precision data from the
AWS logs themselves and avoid the python agent altogether.

Closes scoutapp#631
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants