Hi All,
I have a doubt regarding partitioning of data for every index type.
In our application the data in elastic search is in sync with DB. Each table in DB maps to an index type in elastic search.
We are now creating monthly partitions for all tables in DB, so that the old partitions can dropped as and when required.
Can we create monthly or weekly partitions for all index types in elastic search and drop the partition when its not required?
If not is there an alternative to achieve the partitioning feature in elastic search.
Sounds like you want to have multiple indexes, one per month.
You can do it and have an alias on top of monthly indexes so your front end app will have to query on the alias.
Then, simply remove from the alias the old one and add the new one each month.
Then, drop the old index.
Does it help?
--
David
Twitter : @dadoonet / @elasticsearchfr / @scrutmydocs
Hi All,
I have a doubt regarding partitioning of data for every index type.
In our application the data in Elasticsearch is in sync with DB. Each
table in DB maps to an index type in Elasticsearch.
We are now creating monthly partitions for all tables in DB, so that the
old partitions can dropped as and when required.
Can we create monthly or weekly partitions for all index types in elastic
search and drop the partition when its not required?
If not is there an alternative to achieve the partitioning feature in
Elasticsearch.
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.
