Redis Cluster topology refresh when numbers of errors received from lettuce exceeded set limit

[foxel93]

Hello, I run into an issue when several nodes from our Redis Cluster become unavailable for a period of time, after which they are successfully recovered. At the same moment, client which using lettuce 6.1.2.RELEASE continues getting errors as topology hadn't been refreshed. I tried to enable enableAllAdaptiveRefreshTriggers flag, which didn't help as well.
As I understood, refresh Cluster Topology operation by a set refresh period comes with a huge cost, that's why I can't set a refresh period to 1 minute.
So I am considering using the event of opening circuit breaker to Redis Cluster at our client as the moment for connection recreation.
Is there a chance that I missed something and you have already covered this case?