Filter failed and defunct nodes by default

[srgsanky]

Feature Request

Filter failed and defunct nodes by default while connecting to redis clusters. This is a follow up from #1942

Is your feature request related to a problem? Please describe

It will be good to make this the default behavior in the next major version of lettuce as folks run into temporary errors immediately after a failover. AWS even has a documentation to set the filter https://docs.aws.amazon.com/AmazonElastiCache/latest/red-ug/BestPractices.Clients-lettuce.html#:~:text=Set%20nodeFilter,retrying%20is%20exhausted.

Describe the solution you'd like

Filtering failed/defunct nodes must be the default behavior unless the application want to modify the default for some reason.

Describe alternatives you've considered

A clear and concise description of any alternative solutions or features you've considered.

Teachability, Documentation, Adoption, Migration Strategy

If you can, explain how users will be able to use this and possibly write out a version the docs.
Maybe a screenshot or design?