第五章 MongoDb索引优化 5.1
此章节，是作者阅读网上资料，通过结合自身项目的实践。
引用：原文链接：http://iamcaihuafeng.blog.sohu.com/151638529.html
(本文是一篇转载文章，作者在对MongoDB文档进行了细致的阅读后，总结出了MongoDB的各种索引的用法。)
索引能提高检索数据的速度，你可以想像成在MySQL中创建索引一样，同样索引也是用B-Tree也实现的。
创建schedule的collection：
DB db = m.getDB(
"ms_basic"
);
DBCollection coll = db.getCollection(
"schedule"
);
//插入数据略过
List<
DBObject
> indexList = coll.getIndexInfo();
for
(
DBObject
o : indexList) {
System.
out
.println(
"index ---------"
+ o);
}
运行结果：
index ---------
{ "name" : "_id_" , "ns" : "ms_basic.schedule" , "key" : { "_id" : 1}}
此时可以看到，索引为_id,内置的，按照升序排列
System.
out
.println(
"SELECT * FROM schedule WHERE rsc='ATM033101'"
);
startTime = System.
currentTimeMillis
();
query
=
new
BasicDBObject();
query
.put(
"rsc"
,
"ATM033101"
);
cur = coll.find(
query
);
System.
out
.println(
"查询获得的长度"
+cur.count());
System.
out
.println(
"查询耗
时:"
+(System.
currentTimeMillis
()-startTime)+
"毫秒."
);
运行结果：
-----------华丽分隔线无rsc索引查询---------------
SELECT * FROM schedule WHERE rsc='ATM033101'
查询获得的长度1
查询耗时:63毫秒.
explain分析:{ "cursor" : "BasicCursor" , "indexBounds" : [ ] , "nscanned" : 4597 , "nscannedObjects" : 4597 , "n" : 1 , "millis" : 0 , "oldPlan" : { "cursor" : "BasicCursor" , "indexBounds" : [ ]} , "allPlans" : [ { "cursor" : "BasicCursor" , "indexBounds" : [ ]}]}
看看官方的解释：
public
DBObject
explain
()
Returns an object containing basic information about the exectution of the query that created this cursor This creates a DBObject with the key/value pairs: "cursor" : cursor type "nScanned" : number of records examined by the database for this query "n" : the number of records that the database returned "millis" : how long it took the database to execute the query
1、单列索引
在字段rsc上创建索引，1 (ascending) or -1 (descending)
MongoDb提供两种接口给我们创建索引：
ensureIndex
和
createIndex
具体可以查查在线的API：
http://api.mongodb.org/java/2.0/index.html
coll.createIndex(new BasicDBObject("rsc", 1));
运行：
List<DBObject> indexList =
coll
.getIndexInfo();
for
(DBObject o : indexList) {
System.
out
.println(
"index ---------"
+ o);
}
结果为：
index ---------
{ "name" : "_id_" , "ns" : "ms_basic.schedule" , "key" : { "_id" : 1}}
index ---------
{ "name" : "rsc_1" , "ns" : "ms_basic.schedule" , "key" : { "rsc" : 1}}
再次查找：
query
.put(
"rsc"
,
"ATM033101"
);
-----------华丽分隔线
有
rsc索引查询---------------
SELECT * FROM schedule WHERE rsc='ATM033101'
查询获得的长度1
查询耗时:47毫秒.
explain分析:{ "cursor" : "BtreeCursor rsc_1" , "indexBounds" : [ [ { "rsc" : "ATM033101"} , { "rsc" : "ATM033101"}]] , "nscanned" : 1 , "nscannedObjects" : 1 , "n" : 1 , "millis" : 0 , "oldPlan" : { "cursor" : "BtreeCursor rsc_1" , "indexBounds" : [ [ { "rsc" : "ATM033101"} , { "rsc" : "ATM033101"}]]} , "allPlans" : [ { "cursor" : "BtreeCursor rsc_1" , "indexBounds" : [ [ { "rsc" : "ATM033101"} , { "rsc" : "ATM033101"}]]}]}
此时花费了：
47毫秒
从explain分析中，可以看到：
"cursor" : "BtreeCursor rsc_1" , "indexBounds" : [ [ { "rsc" : "ATM033101"}
，说明使用了索引来进行查询。
2、默认索引
刚刚看到，一共显示有2个索引，包括自己创建的，
其中_id是创建表的时候自动创建的索引，此索引是不能够删除的。
An index is always created on _id. This index is special and cannot be deleted. The _id index enforces uniqueness for its keys.