I tried to run a reindex on a big index, but it failed with the reason of - node is not available.
I seperated the reindex to several reindexes with prefix query on spesific field, so it will go faster.
I can see that the node was up and running all the time, so i dont understand the reason of the failure.
The cluster is big, and the index has 120 shards, with the size of 2.5 tb.
This is the second time it failed. How can i make this more stable?