Hello,
I'm wondering if it's possible to create an elasticsearch index on one node
in the cluster and then when it is complete copy it to another node....(and
ultimately delete from the original). In rare cases it is necessary to
rebuild indexes into smaller indexes. I'm thinking that if we had a
"souped" up server with maybe more memory and/or CPU's we could rebuild the
index on this server without impacting other servers. When the index
re-build is complete we could just tar it up (or copy) it to a different
server...(and maybe just update metadata appropriately).
Hello,
I'm wondering if it's possible to create an elasticsearch index on one
node in the cluster and then when it is complete copy it to another
node....(and ultimately delete from the original). In rare cases it
is necessary to rebuild indexes into smaller indexes. I'm thinking
that if we had a "souped" up server with maybe more memory and/or
CPU's we could rebuild the index on this server without impacting
other servers. When the index re-build is complete we could just tar
it up (or copy) it to a different server...(and maybe just update
metadata appropriately).
Yes you can. In fact the easiest way would be:
start a new node with a different cluster name (so it forms its own
cluster)
index into that node
shut it down
change the cluster name to the main cluster name and restart
-> it joins the main cluster, which imports the new index
and allocates replica shards within the existing nodes
Once cluster is green, shut down the extra node
-> the replica shards are promoted to primaries, and it
allocates new replicas on other machines within the cluster
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.