ElasticArSync
This repository contains easy set up modules for RDB and elasticsearch index with Active Record.
This is based on gem elasticsearch-rails
.
日本語での説明はこちら https://ted-tech.hateblo.jp/entry/2020/08/11/103155
Usage
Preparation
Install gem sidekiq
and gem redis-rails
to use Asynchronous processing for your app.
Install gem elasticsearch-rails
, gem elasticsearch-model
and gem elasticsearch
for your app.
Include module ElasticArSync::Elastic::Syncable
on your Model class which is inherited ActiveRecord.
and define index_config
after include ElasticArSync::Elastic::Syncable
if you use this module normally.
you can override index_config described after.
example below
class XXXXX < ActiveRecord
include ElasticArSync::Elastic::Syncable
index_config
end
then your Model class is included modules which is for sync model's table with elasticsearch index.
when you create new record elasticsearch index save new document in sync. update and delete is same as create.
this module contains class method index_setup
to setup index.
setup means create new index and put alias to new index and sync whole records of RDB with new index, then you can use new index without any command.
XXXXX.index_setup
usable methods
if your Model class included ElasticArSync::Elastic::Syncable
, it can use class methods below.
-
index_setup (class method) → to setup index to elsticsearch
-
create_index (class method) → to create new index to elasticsearch
-
delete_index({index_name}) (class method) → to delete index. needed argument index name which you delete
-
import_all_record(index_name) (class method) → to sync whole record with elasticsearch
-
switch_alias(new_index_name) (class method) → to switch using index, you needed create more than two index
-
mapping_list_keys (class method) → to get mapping list array of your index
-
get_aliases (class method) → to get aliase list array which you created
Customize
if you change process of after commit, you can override instance methods below.
default is provided by ElasticArSync::Elastic::Worker::IndexWorker
-
document_sync_create → for callback of create
-
document_sync_update → for callback of update
-
document_sync_delete → for callback of destroy
if you define original mapping of index like type, attribute, etc,
you can use index_config
after include ElasticArSync::Elastic::Syncable
.
example below
you can override args dynamic
, number_of_shards
, attr_mappings
.
especially default mapping is whole attributes of your Model class, so you can customize mapping attributes by overriding attr_mappings.
index_config(dynamic: 'false', number_of_shards: 1, attr_mappings: { id: 'integer', name: 'text', birth: 'date' })
you can also override part of attributes by override_mappings
like below
*As a premise, your model contain [id: integer, name: string, birth: date]
index_config(override_mappings: { birth: :keyword })
then birth will mapping as keyword type in index of elasticsearch.
if you want add mapping which is not defined as schema column of model, you can save method result as well like below.
index_config(override_mappings: { full_name: :keyword })
*full_name
is instance method
then full_name will mapped to index mapping as keyword type. but you have to define instance method full_name to your model class.
if you define your original mapping additionally, you can also define like below.
settings index: { number_of_shards: number_of_shards } do
mappings do
indexes key, type: value
end
end
tips: just reference here.
https://github.com/elastic/elasticsearch-rails/tree/master/elasticsearch-model
Installation
Add this line to your application's Gemfile:
gem 'elastic_ar_sync'
And then execute:
$ bundle
Or install it yourself as:
$ gem install elastic_ar_sync
Contributing
Contribution directions go here.
License
The gem is available as open source under the terms of the MIT License.