Design Search service using elastic search - rest

Question

I have a requirement. I am building up a search service for a social network. The search service should return the name of the users that somebody searches. Now it will be limited to user domain search only. I am planning to use elastic search to keep the indexes(user domain details). I will then call the EL from my search service(The search service is on nodejs). I am not able to think of a design on how to create the indexes for EL. Should I use a batch to create the indexes or during creation of users I will create the index.
A good pointers or a good design will be appreciated.

Answer 1

you should create index when new user create
simple example may be useful
your user data like:
`
[
{
"_id": "1",
"status": true,
"username": "mak",
"userdomain": "mydomain.com",
"name": "mak doe"
},
{
"_id": "2",
"status": true,
"username": "janny",
"userdomain": "mydomain.com",
"name": "janny"
},
{
"_id": "3",
"status": true,
"username": "mac",
"userdomain": "newdomain.com",
"name": "mac peter"
},
{
"_id": "4",
"status": true,
"username": "mak",
"userdomain": "mydomain.com",
"name": "mak peter"
},
{
"id": "5",
"status": true,
"username": "mak",
"userdomain": "newdomain.com",
"name": "mak peter"
},
]
`
elastic schema look like as below:
`
PUT socialdata
{
"mappings": {
"users": {
"properties": {
"status": {
"type": "boolean"
},
"name": {
"type": "text"
},
"username": {
"type": "string",
"fields": {
"raw": {
"type": "string",
"analyzer": "keyword_lowercase_analyzer"
},
"english": {
"type": "text",
"analyzer": "english"
}
}
},
"userdomain": {
"type": "string",
"fields": {
"raw": {
"type": "string",
"analyzer": "keyword_lowercase_analyzer"
},
"english": {
"type": "text",
"analyzer": "english"
}
}
}
}
}
}
}
`
For Bulk upload :
`
POST socialdata/users/_bulk
{ "index": { "_index": "socialdata","_type": "users", "_id": 1 }}
{"status":true,"username": "mak","userdomain": "mydomain.com","name": "mak doe"}
{ "index": { "_index": "socialdata","_type": "users", "_id": 2 }}
{"status":true,"username": "janny","userdomain": "mydomain.com","name": "janny"}
{ "index": { "_index": "socialdata","_type": "users", "_id": 3 }}
{"status":true,"username": "mac","userdomain": "newdomain.com","name": "mac peter"}
{ "index": { "_index": "socialdata","_type": "users", "_id": 4 }}
{"status":true,"username": "mak","userdomain": "mydomain.com","name": "mak peter"}
{ "index": { "_index": "socialdata","_type": "users", "_id": 5 }}
{"status":true,"username": "mak","userdomain": "newdomain.com","name": "mak peter"}
`
for single index:
`
POST socialdata/users/_bulk
{ "index": { "_index": "socialdata","_type": "users", "_id": 1 }}
{"status":true,"username": "mak","userdomain": "mydomain.com","name": "mak doe"}
`
elastic query :
it will return only two record
`
POST socialdata/users/_search
{
"query": {
"bool": {
"must": [
{
"match": {
"username": "mak"
}
},
{
"match": {
"userdomain": "mydomain.com"
}
}
]
}
}
}
`
it will return only one record
`
POST socialdata/users/_search
{
"query": {
"bool": {
"must": [
{
"match": {
"username": "mak"
}
},
{
"match": {
"userdomain": "newdomain.com"
}
}
]
}
}
}
`