Hi,
Please could someone explain what 'snapshot' indexes are, and what should be done to make my ES cluster be green again? What I've been doing to encounter these is this:
I'm creating a snapshot of an old ElasticSearch cluster that is still running 1.3.4.
I've setup a file system repository and then attempted to create a snapshot with:
curl -XPUT 'http://127.0.0.1:9900/snapshot/es_hot_backup/snapshot_1'
However that gave an error:
{"error":"MapperParsingException[failed to parse, document is empty]","status":400}
Running the command again with 'ignore_unavailable' set, like this:
curl -XPUT "http://127.0.0.1:9900/_snapshot/es_hot_backup/snapshot_1" -d '{
"ignore_unavailable": true
}'
The command appears to have worked
{"accepted":true}
However now the status of the cluster is no longer green, as it appears that a 'snapshot' index has been created, but in a way that means it is not replicated across the cluster:
http://127.0.0.1:9900/_cat/shards?v
index shard prirep state docs store ip node
snapshot 2 p STARTED 0 115b 172.18.0.2 Space Turnip
snapshot 2 r UNASSIGNED
snapshot 0 p STARTED 0 115b 172.18.0.2 Space Turnip
snapshot 0 r UNASSIGNED
snapshot 3 p STARTED 0 115b 172.18.0.2 Space Turnip
snapshot 3 r UNASSIGNED
snapshot 1 p STARTED 0 115b 172.18.0.2 Space Turnip
snapshot 1 r UNASSIGNED
snapshot 4 p STARTED 0 115b 172.18.0.2 Space Turnip
snapshot 4 r UNASSIGNED
(I initially started running the snapshot on production, but switched to a test environment once the initial snapshot command gave a weird error. Yes, I should have done it in test from the start. :p).
Please can someone advise what should I do with that non-replicated index, so as to make the ElasticSearch cluster be green again.
cheers
Dan
Ack