Cut off connections from kube-aggregator to aggregated apiserver lead to panic in the logs #82591
Labels
kind/bug
Categorizes issue or PR as related to a bug.
priority/critical-urgent
Highest priority. Must be actively worked on as someone's top priority right now.
sig/api-machinery
Categorizes an issue or PR as relevant to SIG API Machinery.
Milestone
The kube-aggregator uses Golang's reverse proxy code to forward requests to aggregated API server. Whenever such a connection is uncleanly closed, the Golang code panics. The panic is passed through the handler chain down to the panic handler which prints it.
This leads to main panic stack traces in the kube-apiserver logs whenever an aggregated API server terminates, because every long-running connection (e.g. watches) is uncleanly closed leading to that panic.
This applies to 1.16.
The text was updated successfully, but these errors were encountered: