Add renew slots min interval in cluster to avoid renew slots too frequently when maxAttempts is small

[warrenzhu25]

This is related to #1439. When master node is down and cluster recover by promoting a slave, the Jedis cluster will need as long as connection.timeout * (maxAttempt + 1) to refresh the slots if the master node keep unhealthy. So to shorten the recover time of Jedis cluster, it;s tempted to reduce maxAttempt times. The side effect of this is that sometime when timeout is happening and cluster topology not change, the frequency of refresh is too high. So we want to limit the frequency.

[github-actions[bot]]

This issue is marked stale. It will be closed in 30 days if it is not updated.