Commits on Jun 8, 2022

1. egressIP: node retrieval failure is not respected, causes panic …

   Fixes panic seen during checkEgressNodesReachability
   E0606 11:57:34.063567       1 egressip.go:2096] Node:
   huirwang-0606c-m5w24-worker-2-dkjmc reachability changed, but could not
   retrieve node from cache, err: node
   "huirwang-0606c-m5w24-worker-2-dkjmc" not found
   W0606 11:57:34.063658       1 egressip.go:1657] Unable to remove GARP
   configuration on external logical switch port for egress node:
   huirwang-0606c-m5w24-worker-2-dkjmc, err: object not found
   panic: runtime error: invalid memory address or nil pointer dereference
   [signal SIGSEGV: segmentation violation code=0x1 addr=0x20 pc=0x176626d]
   
   goroutine 487 [running]:
   github.com/ovn-org/ovn-kubernetes/go-controller/pkg/ovn.(*Controller).checkEgressNodesReachability(0xc000171c00)
   	/home/surya/go/src/github.com/openshift/ovn-kubernetes/go-controller/pkg/ovn/egressip.go:2099
   +0x3ed
   created by
   github.com/ovn-org/ovn-kubernetes/go-controller/pkg/ovn.(*Controller).initClusterEgressPolicies
   	/home/surya/go/src/github.com/openshift/ovn-kubernetes/go-controller/pkg/ovn/egressip.go:1772
   +0xdb
   
   The problem here is that the node retrieval fails since the node has
   already been deleted, instead of respecting that failure we just
   log the error saying node object not found and actually continue
   to use that object further in the code which causes this panic.
   
   Signed-off-by: Surya Seetharaman <suryaseetharaman.9@gmail.com>
   (cherry picked from commit 36176e5)

   

   tssurya committed Jun 8, 2022
   Copy the full SHA

   f436d25 View commit details

   Browse the repository at this point in the history