Documentation
¶
Overview ¶
Package cilium provides cluster-level operations against the running Cilium agent DaemonSet (installed by weaver via internal/templates/files/cilium/).
Cluster bootstrap (host-level `cilium install` via the cilium CLI binary) lives in internal/workflows/steps/step_cilium.go and is separate from these helpers, which talk to the running DaemonSet through the kube API.
Index ¶
Constants ¶
const ( // AgentDaemonSetNamespace and AgentDaemonSetName identify the Cilium agent // DaemonSet that weaver installs. AgentDaemonSetNamespace = "kube-system" AgentDaemonSetName = "cilium" // DefaultRolloutTimeout is the conservative deadline for waiting on a Cilium // DaemonSet rolling restart on a small cluster (~30s on a single-node cluster). DefaultRolloutTimeout = 90 * time.Second )
Variables ¶
This section is empty.
Functions ¶
func RestartAgentDaemonSet ¶
func RestartAgentDaemonSet(ctx context.Context, kubeClient *kube.Client, timeout time.Duration) error
RestartAgentDaemonSet triggers a rolling restart of the Cilium agent DaemonSet and blocks until the rollout is complete (or timeout elapses).
This is the documented workaround for the eBPF service reconciler missing Service.spec.type mutations (see hashgraph/solo-weaver issue #619). eBPF programs stay loaded across the restart, so established connections survive; only the agent control plane bounces.
Types ¶
This section is empty.