Problem with Snapshot-Restore in Two-Node-Cluster with Shards


(dbubenheim) #1

Hello guys,

first of all i wanted to say that the Snapshot & Restore functionality is really great for our project. But since we are running two clusters of two nodes each we would like to restore snapshots from one cluster into the other.
We are registering a snapshot repository of type "fs" on one node of the first cluster and and a repository of type "url" on both nodes of the second cluster pointing to the first node of the first cluster.
We do a snapshot of a particular index consisting of four shards and one replica in the first cluster. After that we do a restore on the first node of the second cluster. Everything works fine, but the result is, that only four shards of the restored index are in status STARTED, two are INITIALIZING and two are UNASSINGED.
All in all only the half of the whole index is restored, the other half seems to be missing. Is there anything we can do with the other node of the second cluster? What can we do now?

Thanks in advance. We appreciate any kind of help!
Daniel


(system) #2