My current setup is with 10 nodes with ample space on spinning disks, and 20 nodes with smaller SSD disks. I would like my workflow to be that all data is initially indexed on the SSD nodes, after 10 days is reallocated to the spinning disks, after a further 10 days the index is closed, and after a further 70 days the indexes are deleted. The curator is great for moving them to the spinning disks, but what I am not sure about is how to define that initially all shards (primary and replica) of an index should be created on the ssd nodes. The spinning disk nodes are tagged: node.disk_type: spinning The ssd nodes are tagged node.disk_type: ssd
To transfer after 10 days to spinning: /usr/local/bin/curator --host es101 allocation --older-than 9 --rule disk_type=spinning But how do I define that the default location for all new shards should be on disk_type:ssd ? I have the example here <http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/modules-cluster.html> which I think could be modified like this: curl -XPUT localhost:9200/_cluster/settings -d '{ "persistent" : { "cluster.routing.allocation.require.disk_type" : "ssd" } }' But for one this setting does not exist, and I'm not sure if this will stop the shards being reallocated to spinning later on... Any ideas how to implement my desired workflow? Thank you! -Robin- -- You received this message because you are subscribed to the Google Groups "elasticsearch" group. To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/18615b7c-8e99-42ee-b54d-ef06a1888181%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
