diff --git a/effy/0003/0003_pymongo.py b/effy/0003/0003_pymongo.py
new file mode 100644
index 00000000..7e32a473
--- /dev/null
+++ b/effy/0003/0003_pymongo.py
@@ -0,0 +1,28 @@
+#!/usr/bin/env
+'''
+Instead of Redis, I use MongoDB instead -- since I am using it in 
+a regular basis:)
+Make sure that you have a live mongo connection running on port 27017,
+which is the default port.
+'''
+import uuid
+from pymongo import MongoClient
+## start a client to connect python to mongo
+c = MongoClient()
+## to use a specific db
+## > use mydb 
+db = c['mydb']
+## to create a collection instance to access collection mongoPyPipe
+## I've already created this collection in mongo shell
+collection = db['mongoPyPipe']
+
+## interatively insert doc to collection
+## > db.mongoPyPipe.insert({"id": i, "value": uuid})
+for i in range(200):
+	collection.insert({"id": i, "value": uuid.uuid4()})
+
+
+
+
+
+
diff --git a/effy/0003/readme.txt b/effy/0003/readme.txt
new file mode 100644
index 00000000..89c255df
--- /dev/null
+++ b/effy/0003/readme.txt
@@ -0,0 +1,7 @@
+python和mongo的连接出乎意料的简单。
+需要的只是一个pymongo中的MongoClient来建立一个live connection！
+
+而且最赞的是，由于是nosql，根本不需要考虑collection原来的数据，key，schema什么的，直接统统insert。
+
+在之前和relational sql连接的时候（postgres）需要考虑原来的table里面的PK有没有重复，id要被maintai成城incremental，这种问题。
+
diff --git a/effy/0004/readme.txt b/effy/0004/sampletext.txt
similarity index 100%
rename from effy/0004/readme.txt
rename to effy/0004/sampletext.txt