manager.go: Extend mount timeout to 30 seconds

k8s retries the NodeStageVolume calls wtih an exponential backoff, if we retry for longer internally the wall clock time will be longer until we get hit with a bad backoff (particularly interesting during failovers)

pkg/edgecache/manager.go

@@ -140,7 +140,7 @@ func (m *Manager) callWithConnection(fun ConnectionUsingFunc, endpoint string) e
 
 func (m *Manager) MountVolume(account string, container string, targetPath string) error {
 	return m.callWithConnection(func(conn grpc.ClientConnInterface) error {
-		return sendMount(csi_mounts.NewCSIMountsClient(conn), account, container, targetPath, 500*time.Millisecond, 5*time.Second)
+		return sendMount(csi_mounts.NewCSIMountsClient(conn), account, container, targetPath, 500*time.Millisecond, 30*time.Second)
 	}, m.mountEndpoint)
 }