Here http://www.elasticsearch.org/guide/reference/api/admin-indices-update-settings.html I
read that "Another optimization option is to start the index without any
replicas, and only later adding them". My index is relatively write-heavy,
so when I reindex, I shut down the replica (it's in master=false, data=true
mode) and bring it back later to receive updates in bulk.
Is there any way to make replica "invisible" for the updates, but online
for querying, without shutting it down? And afterwards enable it back so it
will receive updates? In this period it would serve stale data, which is ok.
Here Elasticsearch Platform — Find real-time answers at scale | Elastic I
read that "Another optimization option is to start the index without any
replicas, and only later adding them". My index is relatively write-heavy,
so when I reindex, I shut down the replica (it's in master=false, data=true
mode) and bring it back later to receive updates in bulk.
Is there any way to make replica "invisible" for the updates, but online
for querying, without shutting it down? And afterwards enable it back so it
will receive updates? In this period it would serve stale data, which is ok.
I see. Will it work if I just reboot the server without zen discovery
or with another cluster name so it cannot join the cluster back? I
will try it myself, but maybe there is something I should know
beforehand.
Thanks!
On Sat, Apr 7, 2012 at 11:23 PM, Shay Banon kimchy@gmail.com wrote:
Here Elasticsearch Platform — Find real-time answers at scale | Elastic I
read that "Another optimization option is to start the index without any
replicas, and only later adding them". My index is relatively write-heavy,
so when I reindex, I shut down the replica (it's in master=false, data=true
mode) and bring it back later to receive updates in bulk.
Is there any way to make replica "invisible" for the updates, but online
for querying, without shutting it down? And afterwards enable it back so it
will receive updates? In this period it would serve stale data, which is ok.
Thanks!
--
Regards,
Mikhail Sayapin
["I recommend the art of slow reading."]
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.