fed/clustercontroller: fix race when updating data
updates for ClusterController's maps were made without locking which can lead
to race conditions which were detectede in https://github.com/kubernetes/kubernetes/issues/49958
This change adds a RWMutex to protect the data. We lock and unlock them
whenever the data is accessed to make sure that we dont slow down too much.
Signed-off-by:
Daniel Dao <dqminh89@gmail.com>
Showing
Please
register
or
sign in
to comment