Hi, I am new to Solr. I have a use case to add a new node when an existing node goes down. The new node with a new IP should contain all the replicas that the previous node had. So I am using a network storage (cinder storage) in which the data directory (where the solr.xml and the core directories resides) is getting created when a node starts up. The new node with a new IP after the replacement will contain the same set of directories which the old node had. I have noticed the new node is added to the cluster without the need for an ADDREPLICA.
Is this an expected behavior in Solr? Does ZK still hold the references to old node? What's the recommended solution if I want to re-use the data directory associated with the old new while spinning up a new node. The goal is to avoid data loss and to reduce the time taken to recover a node. Thanks in advance! -Minu
