github/kubernetes
1
0
Fork 0
mirror of https://github.com/k3s-io/kubernetes.git synced 2025-07-26 05:03:09 +00:00
Code Issues Projects Releases Wiki Activity

Merge pull request #58186 from negz/master

Browse Source 
Automatic merge from submit-queue (batch tested with PRs 57266, 58187, 58186, 46245, 56509). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.

Avoid panic in Cloud CIDR Allocator

**What this PR does / why we need it**:
I suspect a race exists where we attempt to look up the CIDR for a terminating node. By the time `updateCIDRAllocation` is called the node has disappeared. We determine it does not have a cloud CIDR (i.e. Alias IP Range) and attempt to record a `CIDRNotAvailable` node status. Unfortunately we reference `node.Name` while `node` is still nil.

By getting the node before looking up the cloud CIDR we avoid the nil pointer dereference, and potentially fail fast in the case the node has disappeared.

**Which issue(s) this PR fixes**:
Fixes #58181

**Release note**:

```release-note
Avoid panic when failing to allocate a Cloud CIDR (aka GCE Alias IP Range). 
```
This commit is contained in:
Kubernetes Submit Queue 2018-01-12 20:34:50 -08:00 committed by GitHub
commit 15ef3a8ba4
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 6 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
pkg/controller/nodeipam/ipam/cloud_cidr_allocator.go
View File

@ -191,10 +191,14 @@ func (ca *cloudCIDRAllocator) AllocateOrOccupyCIDR(node *v1.Node) error {
// updateCIDRAllocation assigns CIDR to Node and sends an update to the API server. // updateCIDRAllocation assigns CIDR to Node and sends an update to the API server.
func (ca *cloudCIDRAllocator) updateCIDRAllocation(nodeName string) error { func (ca *cloudCIDRAllocator) updateCIDRAllocation(nodeName string) error {
var err error
var node *v1.Node
defer ca.removeNodeFromProcessing(nodeName) defer ca.removeNodeFromProcessing(nodeName)
node, err := ca.nodeLister.Get(nodeName)
if err != nil {
glog.Errorf("Failed while getting node %v for updating Node.Spec.PodCIDR: %v", nodeName, err)
return err
}
cidrs, err := ca.cloud.AliasRanges(types.NodeName(nodeName)) cidrs, err := ca.cloud.AliasRanges(types.NodeName(nodeName))
if err != nil { if err != nil {
nodeutil.RecordNodeStatusChange(ca.recorder, node, "CIDRNotAvailable") nodeutil.RecordNodeStatusChange(ca.recorder, node, "CIDRNotAvailable")
@ -210,12 +214,6 @@ func (ca *cloudCIDRAllocator) updateCIDRAllocation(nodeName string) error {
} }
podCIDR := cidr.String() podCIDR := cidr.String()
node, err = ca.nodeLister.Get(nodeName)
if err != nil {
glog.Errorf("Failed while getting node %v for updating Node.Spec.PodCIDR: %v", nodeName, err)
return err
}
if node.Spec.PodCIDR == podCIDR { if node.Spec.PodCIDR == podCIDR {
glog.V(4).Infof("Node %v already has allocated CIDR %v. It matches the proposed one.", node.Name, podCIDR) glog.V(4).Infof("Node %v already has allocated CIDR %v. It matches the proposed one.", node.Name, podCIDR)
// We don't return here, in order to set the NetworkUnavailable condition later below. // We don't return here, in order to set the NetworkUnavailable condition later below.