#Introduction

This is the readme file for Udacity Logs Analysis Project

Set up vagrant

1. Go virtualbox.org

2. Download and install virtualbox for your operating systems

3. Go vagrantup.com

4. Download and install for your operating systems

5. Clone fullstack-nanodegree-vm

6. Change directory to vagrant folder

7. Run this code to start vagrant

   $ vagrant up
$ vagrant ssh

##Getting Started

1. Open logAnalysis folder

2. Run psql -d news -f newsdata.sql in terminal to connect to the news database and run newsdata.sql

3. Run python newsdb.py to setup news database query

   • Result of the query should appear in the terminal after this

4. Run python logAnalysis to Run the site

5. Browse the site at http://0.0.0.0:8000/

##Question 1 (Top Article) Top Article Link

select articles.title as title, count(articles.title) as views        
from articles inner join log on (log.path='/article/'||articles.slug)            
group by title        
order by views           
desc     
limit 3;     

##Question 2 (Top Author) Top Author Link

select authors.name as name, count(authors.name) as views  
from authors right outer join (articles inner join log on (log.path='/article/'||articles.slug)) as ntable  
on (authors.id=ntable.author)  
group by name  
order by views  
desc;

##Question 3 (On which days did more than 1% of requests lead to errors? ) Most Errors Link

1. Create a view to organize the data

create view count_table  
 as select date(time), count(status)  
 filter (where status='404 NOT FOUND') as nfcount,  
 count(time) as total_request  
 from log  
 group by date(time);

2. Get the result by entering this code

select date,  
 round( cast(float8 (nfcount*100::decimal)/total_request as numeric),2) as error_percentage  
 from count_table  
 where round( cast(float8 (nfcount*100::decimal)/total_request as numeric),2) > 1