You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
UPSTREAM: <carry>: apiserver: log new connections during termination
UPSTREAM: <carry>: apiserver: create LateConnections events on events in the last 20% of graceful termination time
UPSTREAM: <carry>: apiserver: log source in LateConnections event
UPSTREAM: <carry>: apiserver: skip local IPs and probes for LateConnections
UPSTREAM: <carry>: only create valid LateConnections/GracefulTermination events
UPSTREAM: <carry>: kube-apiserver: log non-probe requests before ready
UPSTREAM: <carry>: apiserver: create hasBeenReadyCh channel
UPSTREAM: <carry>: kube-apiserver: log non-probe requests before ready
UPSTREAM: <carry>: kube-apiserver: log non-probe requests before ready
UPSTREAM: <carry>: fix termination event(s) validation failures
UPSTREAM: <carry>: during the rebase collapse to create termination event
it makes recording termination events a non-blocking operation.
previously closing delayedStopCh might have been delayed on preserving data in the storage.
the delayedStopCh is important as it signals the HTTP server to start the shutdown procedure.
it also sets a hard timeout of 3 seconds for the storage layer since we are bypassing the API layer.
UPSTREAM: <carry>: rename termination events to use lifecycleSignals
OpenShift-Rebase-Source: 15b2d2e
UPSTREAM: <carry>: extend termination events
- we tie the shutdown events with the UID of the first
(shutdown initiated), this provides us with a more
deterministic way to compute shutdown duration from
these events
- move code snippets from the upstream file to openshift
specific patch file, it reduces chance of code conflict
0 commit comments