Can someone let me know what rules do elasticsearch follows to map a string
field to text
/keyword
when an elasticsearch cluster running 2.3 is upgraded to 5.6.16?
For example how will the below string properties map to text/keyword?
{
"properties": {
"name" : {
"type" : "string",
"fields" : {
"lower" : {
"type" : "string",
"analyzer" : "raw_lowercase_analyzer"
},
"raw" : {
"type" : "string",
"index" : "not_analyzed",
"ignore_above" : 10922,
"eager_global_ordinals": true
}
}
}
}
}```