Some of these objects use a lot of lazy loading so only part of their
properties may be loaded. When they need to be saved, if I index them, some
of the properties will be juste removed from the index if they have not been
loaded and loaded them again is a pain in the ... for performances.
Some of these objects use a lot of lazy loading so only part of their
properties may be loaded. When they need to be saved, if I index them, some
of the properties will be juste removed from the index if they have not been
loaded and loaded them again is a pain in the ... for performances.
Some of these objects use a lot of lazy loading so only part of their
properties may be loaded. When they need to be saved, if I index them, some
of the properties will be juste removed from the index if they have not been
loaded and loaded them again is a pain in the ... for performances.
I've pushed preliminary ground work to support update endpoint (that will
still mean reindexing the docs). And, one of elasticsearch users is helping
with actually implementing an Update endpoint (it will start with "update by
script"). The initial solution will be update a single doc, not bulk update.
Some of these objects use a lot of lazy loading so only part of their
properties may be loaded. When they need to be saved, if I index them,
some
of the properties will be juste removed from the index if they have not
been
loaded and loaded them again is a pain in the ... for performances.
This is a derivation of the partial update plugin. It was more or less a
thought experiment on building a _reindex endpoint that could (internally)
migrate indices. I'd like to see webhooks introduced into the bulk APIs so
the caller could just register callback methods (HTTP POST) similar to
jQuery (onSuccess, onFailure, onComplete).
I tested this lightly with 100K small-ish records and reindexing was, of
course, faster than the original index times since everything is handled on
the server side. Ideally, this plugin would accept a query to run. At the
moment it just uses a match match_all query.
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.