diff --git a/Readme.md b/Readme.md
index 5eacb3c..dbb28c7 100644
--- a/Readme.md
+++ b/Readme.md
@@ -1,6 +1,6 @@
 
 # SQL chart
-Draw chart any SQL queries
+Draw chart for any SQL queries
 
 
 ## What is this?
@@ -13,7 +13,17 @@ A simple web application to visualize data from any arbitrary SQL query.
 * available columns in the data will be shown in UI. visualize the data by dragging columns into required axis. Also choose requied aggregation function for pivot table
 * Entire state of the graph ( SQL, columns in each axis, aggregation function etc ) is stored in the URL ( in URL hash ) so that we can reload the page without loosing the data.
 
+## How to run this
+```bash
+git clone https://github.com/harish2704/sql-chart
+cd sql-chart
+pip2 install -r requirements.txt --user
+# Edit config file and set database connection url. For details, refer https://docs.sqlalchemy.org/en/latest/core/engines.html
+./run.sh
+```
+
+
 ## Thanks
 * Authors of pivottable ( https://github.com/nicolaskruchten/pivottable ). The UI code is shamelessly copied from one of the examples given in this project.
 * Simple backed api server ( server.js ) is written in https://falconframework.org/.
-* https://plot.ly/javascript/
\ No newline at end of file
+* https://plot.ly/javascript/
diff --git a/run.sh b/run.sh
index ee83456..c1ce1a1 100755
--- a/run.sh
+++ b/run.sh
@@ -1,4 +1,4 @@
 #!/usr/bin/env bash
 
-[ -f ./env.sh ] && . ./env.sh && export DB_CONNECTION
+[ -f ./config ] && . ./config && export DB_CONNECTION
 gunicorn server:api
diff --git a/server.py b/server.py
index 8516162..d871d05 100644
--- a/server.py
+++ b/server.py
@@ -6,7 +6,7 @@ import datetime
 import os
 
 db_url = os.environ[ 'DB_CONNECTION' ] if 'DB_CONNECTION' in os.environ else 'postgresql://postgres@localhost/postgres'
-eng = sqlalchemy.create_engine(db_url, pool_size=1, pool_recycle=3600)
+eng = sqlalchemy.create_engine(db_url, pool_recycle=3600)
 conn = eng.connect()
 print "Connected to db"