I am trying to find the best practice to update a document with nested objects: I am using the ES as a view cache:
let say I have tow entities in my SQL DB: the First one is Address and the Homes and the relation between them is one too many
Address Table has columns=> Id , Street, ZipCode
Home Table has colmns=> Id, AddressId(FK), price
And I store them in ES aggregated:
Home : {
Id: 1,
price : 3000,
Address :{
id: 1,
street: "blabla",
zibcode : 12345
}
}
if I made any change on Address I want to update all Home Documents that contains the same Address in the Home index with the new value. What I need to know how can I update all Documents in the same index in one request to ES.
_update_by_query
is probably what you're after.
This ain't pretty but it should do the job:
PUT test/_doc/1
{
"Id": 1,
"price": 3000,
"Address": {
"id": 1,
"street": "blabla",
"zipcode": 12345
}
}
POST test/_update_by_query
{
"query": {
"term": {
"Address.id": 1
}
},
"script": {
"source": "ctx._source['Address']['id'] = 2;ctx._source['Address']['street'] = 'foo';ctx._source['Address']['zipcode'] = 5678"
}
}
GET test/_doc/1
And the output then is:
{
"_index" : "test",
"_type" : "_doc",
"_id" : "1",
"_version" : 9,
"_seq_no" : 8,
"_primary_term" : 1,
"found" : true,
"_source" : {
"Address" : {
"zipcode" : 5678,
"street" : "foo",
"id" : 2
},
"price" : 3000,
"Id" : 1
}
}