If we have a stretch kafka cluster of two DCs with 6 brokers and 5 zookeepers, split in below fashion:
DC1 : 3 zookeepers & 3 brokers ; DC2 : 2 zookeepers & 3 brokers
For a DR scenario in case, DC1 fails. How can we use a spare zookeeper, say an observer zookeeper(preferably in cloud) to come for the rescue and avoid a disaster? Can an extra zookeeper serve a DR situation? Suggestions welcome!
Thanks!