indices_boost seems to have no effect

[clintongormley]

Hiya

Maybe I'm testing the wrong thing, but it doesn't look like indices_boost is having any effect:

Load this data:

curl -XPUT 'http://127.0.0.2:9200/es_test_1/'  -d '
{}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_2/'  -d '
{}
'
curl -XPUT 'http://127.0.0.2:9200/_all/type_1/_mapping'  -d '
{
   "_all" : {
      "store" : "yes",
      "term_vector" : "with_positions_offsets"
   },
   "properties" : {
      "num" : {
         "store" : "yes",
         "type" : "integer"
      },
      "date" : {
         "format" : "yyyy-MM-dd HH:mm:ss",
         "type" : "date"
      },
      "text" : {
         "store" : "yes",
         "type" : "string"
      }
   }
}
'
curl -XPUT 'http://127.0.0.2:9200/_all/type_2/_mapping'  -d '
{
   "_all" : {
      "store" : "yes",
      "term_vector" : "with_positions_offsets"
   },
   "properties" : {
      "num" : {
         "store" : "yes",
         "type" : "integer"
      },
      "date" : {
         "format" : "yyyy-MM-dd HH:mm:ss",
         "type" : "date"
      },
      "text" : {
         "store" : "yes",
         "type" : "string"
      }
   }
}
'
curl -XPOST 'http://127.0.0.2:9200/_refresh' 
curl -XGET 'http://127.0.0.2:9200/_cluster/health?timeout=2s&wait_for_status=green' 
curl -XPUT 'http://127.0.0.2:9200/es_test_1/type_1/1/_create'  -d '
{
   "num" : "2",
   "date" : "2010-04-2 00:00:00",
   "text" : "foo"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_1/type_2/2/_create'  -d '
{
   "num" : "3",
   "date" : "2010-04-3 00:00:00",
   "text" : "foo"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_2/type_1/3/_create'  -d '
{
   "num" : "4",
   "date" : "2010-04-4 00:00:00",
   "text" : "foo"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_2/type_2/4/_create'  -d '
{
   "num" : "5",
   "date" : "2010-04-5 00:00:00",
   "text" : "foo"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_1/type_1/5/_create'  -d '
{
   "num" : "6",
   "date" : "2010-04-6 00:00:00",
   "text" : "foo bar"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_1/type_2/6/_create'  -d '
{
   "num" : "7",
   "date" : "2010-04-7 00:00:00",
   "text" : "foo bar"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_2/type_1/7/_create'  -d '
{
   "num" : "8",
   "date" : "2010-04-8 00:00:00",
   "text" : "foo bar"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_2/type_2/8/_create'  -d '
{
   "num" : "9",
   "date" : "2010-04-9 00:00:00",
   "text" : "foo bar"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_1/type_1/9/_create'  -d '
{
   "num" : "10",
   "date" : "2010-04-10 00:00:00",
   "text" : "foo bar baz"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_1/type_2/10/_create'  -d '
{
   "num" : "11",
   "date" : "2010-04-11 00:00:00",
   "text" : "foo bar baz"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_2/type_1/11/_create'  -d '
{
   "num" : "12",
   "date" : "2010-04-12 00:00:00",
   "text" : "foo bar baz"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_2/type_2/12/_create'  -d '
{
   "num" : "13",
   "date" : "2010-04-13 00:00:00",
   "text" : "foo bar baz"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_1/type_1/13/_create'  -d '
{
   "num" : "14",
   "date" : "2010-04-14 00:00:00",
   "text" : "bar baz"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_1/type_2/14/_create'  -d '
{
   "num" : "15",
   "date" : "2010-04-15 00:00:00",
   "text" : "bar baz"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_2/type_1/15/_create'  -d '
{
   "num" : "16",
   "date" : "2010-04-16 00:00:00",
   "text" : "bar baz"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_2/type_2/16/_create'  -d '
{
   "num" : "17",
   "date" : "2010-04-17 00:00:00",
   "text" : "bar baz"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_1/type_1/17/_create'  -d '
{
   "num" : "18",
   "date" : "2010-04-18 00:00:00",
   "text" : "baz"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_1/type_2/18/_create'  -d '
{
   "num" : "19",
   "date" : "2010-04-19 00:00:00",
   "text" : "baz"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_2/type_1/19/_create'  -d '
{
   "num" : "20",
   "date" : "2010-04-20 00:00:00",
   "text" : "baz"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_2/type_2/20/_create'  -d '
{
   "num" : "21",
   "date" : "2010-04-21 00:00:00",
   "text" : "baz"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_1/type_1/21/_create'  -d '
{
   "num" : "22",
   "date" : "2010-04-22 00:00:00",
   "text" : "bar"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_1/type_2/22/_create'  -d '
{
   "num" : "23",
   "date" : "2010-04-23 00:00:00",
   "text" : "bar"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_2/type_1/23/_create'  -d '
{
   "num" : "24",
   "date" : "2010-04-24 00:00:00",
   "text" : "bar"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_2/type_2/24/_create'  -d '
{
   "num" : "25",
   "date" : "2010-04-25 00:00:00",
   "text" : "bar"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_1/type_1/25/_create'  -d '
{
   "num" : "26",
   "date" : "2010-04-26 00:00:00",
   "text" : "foo baz"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_1/type_2/26/_create'  -d '
{
   "num" : "27",
   "date" : "2010-04-27 00:00:00",
   "text" : "foo baz"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_2/type_1/27/_create'  -d '
{
   "num" : "28",
   "date" : "2010-04-28 00:00:00",
   "text" : "foo baz"
}
'
curl -XPUT 'http://127.0.0.2:9200/es_test_2/type_2/28/_create'  -d '
{
   "num" : "29",
   "date" : "2010-04-29 00:00:00",
   "text" : "foo baz"
}
'
curl -XPOST 'http://127.0.0.2:9200/_refresh' 
curl -XGET 'http://127.0.0.2:9200/_cluster/health?timeout=2s&wait_for_status=green' 

Then try these queries:

curl -XGET 'http://127.0.0.2:9200/_all/_search?pretty=true'  -d '
{
   "sort" : [
      "score"
   ],
   "fields" : [],
   "query" : {
      "field" : {
         "text" : "foo"
      }
   },
   "explain" : 0,
   "indices_boost" : {
      "es_test_2" : 0.1,
      "es_test_1" : 5000
   },
   "size" : 2
}
'

and

curl -XGET 'http://127.0.0.2:9200/_all/_search?pretty=true'  -d '
{
   "sort" : [
      "score"
   ],
   "fields" : [],
   "query" : {
      "field" : {
         "text" : "foo"
      }
   },
   "explain" : 0,
   "indices_boost" : {
      "es_test_2" : 5000,
      "es_test_1" : 0.1
   },
   "size" : 2
}
'

They both return:

# {
#    "hits" : {
#       "hits" : [
#          {
#             "_index" : "es_test_1",
#             "_id" : "2",
#             "_type" : "type_2"
#          },
#          {
#             "_index" : "es_test_2",
#             "_id" : "4",
#             "_type" : "type_2"
#          }
#       ],
#       "total" : 16
#    },
#    "_shards" : {
#       "failed" : 0,
#       "successful" : 10,
#       "total" : 10
#    }
# }

I would have thought that they would have returned the same docs, but in the opposite order. Instead, they both return the docs in the same order: (without the fields param):

# {
#    "hits" : {
#       "hits" : [
#          {
#             "_source" : {
#                "num" : "3",
#                "text" : "foo",
#                "date" : "2010-04-3 00:00:00"
#             },
#             "_index" : "es_test_1",
#             "_id" : "2",
#             "_type" : "type_2"
#          },
#          {
#             "_source" : {
#                "num" : "5",
#                "text" : "foo",
#                "date" : "2010-04-5 00:00:00"
#             },
#             "_index" : "es_test_2",
#             "_id" : "4",
#             "_type" : "type_2"
#          }
#       ],
#       "total" : 16
#    },
#    "_shards" : {
#       "failed" : 0,
#       "successful" : 10,
#       "total" : 10
#    }
# }

When i do an explain, it seems to be taking the boost into account, at least for one index:

curl -XGET 'http://127.0.0.2:9200/_all/_search?pretty=true'  -d '
{
   "sort" : [
      "score"
   ],
   "fields" : [],
   "query" : {
      "field" : {
         "text" : "foo"
      }
   },
   "explain" : 1,
   "indices_boost" : {
      "es_test_2" : 5000,
      "es_test_1" : 0.1
   },
   "size" : 2
}
'
# {
#    "hits" : {
#       "hits" : [
#          {
#             "_index" : "es_test_1",
#             "_id" : "2",
#             "_type" : "type_2",
#             "_explanation" : {
#                "value" : 1.287682,
#                "details" : [
#                   {
#                      "value" : 1,
#                      "description" : "tf(termFreq(text:foo)=1)"
#                   },
#                   {
#                      "value" : 1.287682,
#                      "description" : "idf(docFreq=2, maxDocs=4)"
#                   },
#                   {
#                      "value" : 1,
#                      "description" : "fieldNorm(field=text, doc=0)"
#                   }
#                ],
#                "description" : "fieldWeight(text:foo in 0), product of:"
#             }
#          },
#          {
#             "_index" : "es_test_2",
#             "_id" : "4",
#             "_type" : "type_2",
#             "_explanation" : {
#                "value" : 1.1823214,
#                "details" : [
#                   {
#                      "value" : 0.99999994,
#                      "details" : [
#                         {
#                            "value" : 5000,
#                            "description" : "boost"
#                         },
#                         {
#                            "value" : 1.1823215,
#                            "description" : "idf(docFreq=4, maxDocs=6)"
#                         },
#                         {
#                            "value" : 0.00016915871,
#                            "description" : "queryNorm"
#                         }
#                      ],
#                      "description" : "queryWeight(text:foo^5000.0), product of:"
#                   },
#                   {
#                      "value" : 1.1823215,
#                      "details" : [
#                         {
#                            "value" : 1,
#                            "description" : "tf(termFreq(text:foo)=1)"
#                         },
#                         {
#                            "value" : 1.1823215,
#                            "description" : "idf(docFreq=4, maxDocs=6)"
#                         },
#                         {
#                            "value" : 1,
#                            "description" : "fieldNorm(field=text,doc=0)"
#                         }
#                      ],
#                      "description" : "fieldWeight(text:foo in 0),product of:"
#                   }
#                ],
#                "description" : "weight(text:foo^5000.0 in 0), product of:"
#             }
#          }
#       ],
#       "total" : 16
#    },
#    "_shards" : {
#       "failed" : 0,
#       "successful" : 10,
#       "total" : 10
#    }
# }

[kimchy]

I fixed it just now in #143, and now the boost per index is actually a boost factor (the explanation reflects that as well...). Can you verify that it works for you?

[kimchy]

ok, closing this, it should work (have tests). If not, reopen...