Skip to content
This repository has been archived by the owner on Sep 13, 2022. It is now read-only.

Spans being dropped by udp_sender #206

Closed
black-adder opened this issue Jan 23, 2018 · 3 comments
Closed

Spans being dropped by udp_sender #206

black-adder opened this issue Jan 23, 2018 · 3 comments
Assignees
@black-adder
Labels
bug

Comments

@black-adder
Copy link
Contributor

I wrote a load test that creates 100 traces at once. The udp_sender seemingly successfully sends the batches to agent with no errors. However, I check the UI and I don't see all 100 traces. So I used tcpdump to see how many packets were arriving at the agent and indeed some packets were missing.

Given the fire and forget nature of udp, I guess this is valid? However, the dropped spans aren't observable since the dgram udp socket never returns an error even if the packet is dropped.
@black-adder black-adder self-assigned this Jan 23, 2018
@black-adder
Copy link
Contributor Author

@tiffon you know anything we could do about this?

@MarckK MarckK mentioned this issue Jan 23, 2018
Closed
@tiffon
Copy link
Member

tiffon commented Jan 23, 2018

@black-adder How is this handled in the other clients? As long as the approach is feasible and practical in node, we should aim for the same approach, or similar to be consistent.

@black-adder
Copy link
Contributor Author

sigghhhhh, it was #124

Iuriy-Budnikov pushed a commit to agile-pm/jaeger-client-node that referenced this issue Sep 25, 2021
@mayurkale22
fix(basic-tracer): performance.now() is relative from start time (jae…
c2a7feb 
…gertracing#206)

* fix(basic-tracer): performance.now() is relative from start time

* refactor(time): tuples

* refactor(span): change name of toHrTime

* fix(span): duration default

* fix(hrtime): conver to tuple

* fix(basic-tracer): nanosecond accuracy

* feat(core): move time utils to core

* feat(core): add millis and nanos converter

* test(basic-tracer): fix span time tests
Iuriy-Budnikov pushed a commit to agile-pm/jaeger-client-node that referenced this issue Sep 25, 2021
@mayurkale22
Continuation of jaegertracing#206 (jaegertracing#279)
7fff538 
* fix: performance.now() is relative from start time

* fix: grpc plugin tests

* fix: import performance
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Assignees

@black-adder black-adder

Labels
bug
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@black-adder @tiffon