-
Notifications
You must be signed in to change notification settings - Fork 1.5k
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
If you open an OTLP receiver in port 4317, another one is opened on port 55680 #2565
Comments
BTW, this was done intentionally as part of #2104 for the transition period |
@bogdandrutu @bhautikpip submitted a PR to remove support for port 55680. If we decide to deprecate this port, we should do this before GA. Please advise. cc: @mxiamxia can you please take a look based on what we hear from @bogdandrutu |
@alolita I think we need to agree on the right thing to do here. probably we should discuss this during SIG meeting to achieve agreement. The first goal is to make a decision. To understand: "If we release with this it means we have to always support this behavior" |
Action Item: We should scan and remove all references in the code and docs to the old port before GA. If we find a reference, file an issue, we will need to fix it. See: https://github.com/search?q=org%3Aopen-telemetry+55680&type=code |
Removed all the legacy port(55680) mentions across the OpenTelemetry repos. The port will be disabled in OTel Collector v0.30.0 per the discussion below. |
This is completed from an implementation standpoint so I will close this issue. I will open another issue to track user feedback until the Collector reaches version 0.30.0 when this port will be disabled. |
opentelemetry-collector/receiver/otlpreceiver/otlp.go
Line 129 in e6319ac
Currently if you open an OTLP receiver on port 4317, the code is wired to open an extra one in 55680.
I guess that this was meant as a legacy or backwards compatible feature, but it can trip setups that trying to use 55680 for other uses (it tripped me up while I was doing a proof of concept).
In general I think this is very unintuitive behaviour that is rather tricky to debug.
The text was updated successfully, but these errors were encountered: