Indexing.txt

Indexing 

1. Collectiion Scan(Without Indexing)


Type OF INDEX 

1 . Single Feild Indexes 
2 . Compound Feild Indexes
3 . Text Indexes

1 . Single Feild Index

    Using single feild index will be created 
    db.collection.getindexes()  ==> Give lost of Index for collection
    db.collection.createdIndex({feild : 1})  ===> return name or object
    db.dropIndex() ==> Drop Index


2 . Compund Scan 
                     Merge two or more feils to create an Index 
    ex : db,collection.createdIndex({username : 1 , rid : 1})
        
        will create index named username_1_rid_1 . 

        Now if we query for any document like 
           db.collection.find({username : "xyz , rid: "23u2o492"}).explain('executinStats')

           then it only "totalDocsExamined" will be only  1 


           if query is something like 
           db.collection.find({username : "xy"}).explain('executinStats')

           then even result is same , but for below query result might get different

           db.collection.find({rid: "23u2o492"}).explain('executinStats')


           for this query "totalDocsExamined" will be only  larger or we can say max , and it user "CALLSCAN" Collection sacn for identifying document

           and other query users "IndexScan" for serching 



           ===>If yoy want index to be applied on unique value we can set it as 
               
               db.collection.createIndex({name : 1} , {unique : true})

               prevent same name value by sorting B tree in asc order by name feild 



        ===>  we even can set filter inside filter 


        db.collection.createdIndex({age : 1 , {partialFilterExression : {age : {$gt : 22}}}})


        Now only documents with age gt then 22 will be added in index tree / B tree 

        NOw if i run query like ==> db.collection.find({age : {$lt : 21}})

        Now for this query CALLSCAN will happen , 

        if we set our filter for age >= 22 then and then INDEXSCAN (index scan ) will be triggered 


        Exmaple : 2

        db.collection.createdIndex({name : 1 , {partialFilterExression : {rid : {$exists : true}}}})


        take only those documents in inside index tree which has rid feild , documents which has no existance of rid feils are not been inculded inside tree 




        =======> Covered Query 

        All the feilds in query are the part of an index 
        All the feilds returned in the query are in the same index 

        Covered query are very fast queries ,let us understad it by example 

        db.collection.createdIndex({age : 1 , gender : 1})

        Index is been made of a combinatin of age and gender 

        now covered query is 

        db.collection.find({age : {$gt : 21} , gender : "male"})



        
        =====> Winnig plan 

        if for one query there is more then two or more then one index is available then before triggereing the query , or i don't know the exact time but ya mongo messure the perfomance for both index for a sample  collection and caches ihe index for a while(almost for next 1000 write operation) , and used it for other simillar operation 



        ====> Multi Key Index 

        used on array and its feilds(Arrays value)

        It creates index for array and for each value of array too .



        ===> Text base Index 

            ==> Query for create Text base Index 

        db['user-as-individual'].createIndex({firstname : "text" , lastname : "text" , username : "text"} , {weights : {firstname : 2, lastname : 1 , username : 2}})

        ===> Retrive data bases on text 

         db['user-as-individual'].find({$text : {$search : "harsh official"}})


        ====>   Add score for each result 
          db['user-as-individual'].find({$text : {$search : "harsh official"}}, {hamaraScore }).pretty()