When I run a python script in a terminal it runs as expected; downloads file and saves it in the desired spot.

sudo python script.py 

I've added the python script to the root crontab, but then it runs as it is supposed to except it does not write the file.

$ sudo crontab -l
> * * * * * python /home/test/script.py >> /var/log/test.log 2>&1

Below is a simplified script that still has the problem:

#!/usr/bin/python
scheduleUrl = '
schedule = '/var/test/schedule.xml'
# Download url and save as filename
def wget(url, filename): import urllib2 try: response = urllib2.urlopen(url) except Exception: import traceback logging.exception('generic exception: ' + traceback.format_exc()) else: print('writing:'+filename+';') output = open(filename,'wb') output.write(response.read()) output.close()
# Download the schedule
wget(scheduleUrl, schedule)

I do get the message "writing:name of file;" inside the log, to which the cron entry outputs. But the actual file is nowhere to be found...

The dir /var/test is chmodded to 777 and using whatever user, I am allowed to add and change files as I please.

4 Answers

• Check log files grep -i cron /var/log/syslog
• Add an empty line to the end of the crontab, This has been a known bug for ages, not sure if it is solved.
• Remove the 2>&1 from the command line until it works as designed. Any usefull errors are redirected to a file that is not created; effectively lost.
• Check if root received mail (eg. using mutt or in /var/spool/mail). Error messages from cron are sent to system email by default.

Also:

• Reconsider the 777 permissions as soon as possible. When running from root, 755 root:root should be sufficient to be able to check the logs from unprivileged user)
• Reconsider running the script from root, it is bad practise.

What works for me

Crontab

#Borrowed from anacron
SHELL=/bin/sh
PATH=/usr/local/sbin:/usr/local/bin:/sbin:/bin:/usr/sbin:/usr/bin
#End borrowed from anacron
* * * * * python /home/username/somedir/test.py

Python script

scheduleUrl = '
schedule = '/tmp/test.html'
# Download url and save as filename
def wget(url, filename): import urllib2 try: response = urllib2.urlopen(url) except Exception: import logging logging.exception('error') else: print('writing:'+filename+';') output = open(filename,'wb') output.write(response.read()) output.close()
# Download the schedule
wget(scheduleUrl, schedule)

Added environment variables. Used tmp instead of var to verify there weren't permissions issues.

I had a similar problem:

f = open('./my_file.txt', 'w')
f.close()

Was not opening and writing the file when run from cron. This solved it

f = open('<full_path_of_file>/my_file.txt', 'w')
f.close()

For me the solution was as simple as changing the file access mode. Instead of:

output = open(filename,'wb')

Try:

output = open(filename,'rb+')

I used this to scrape craigslist for job postings (for myself) and to model into a database. All done on a raspberry pi.

I found this and it's very similar to the issue I was having, though I didn't quite get my answer from here. The cron job was executing but the python script would not write files when executed via a cron job. The script would write web scraped text files if executed from the command line.

Solution for me was simply the wb to rb+rb+ Opens a file for both reading and writing in binary format. The file pointer placed at the beginning of the file.