I am in the process of switching from Google Analytics to Piwik and would like to import all historic data from my Apache log files. To that end I have concatenated all available logs and placed them in the import_logs folder with the filename historic.logs. However, what I do next is not at all clear. The Piwik documentation suggests something along the lines of
python /var/www/html/piwik/misc/log-analytics/import_logs.py --idsite=example.net --url=http://example.net historic.logs
This makes little sense to me. Why do I need a URL of any kind if I already have the logs available locally? What precisely is the site id?
I hope that someone here will be able to help.
It would be nice if we could have a more user-friendly feature, inside the Piwik control panel (perhaps a plugin?), that could import apache logs from a specified folder inside the Piwik installation.
Then the user would only need to a) copy new apache logs to the right folder or b) set up a very simple cron job to copy new apache logs once a day in the folder.
Matt, there are users who have shared hosting accounts without the ability to run python. They can only use cron jobs.
FredAt: --idsite and --url are is the site ID and URL for the website you are tracking that is configured on your piwik instance. Edit: the URL is that of the piwik installation.
Go to the admin page
You may have more than one site configured.
PeterV: there are extremely cheap shared hosting services that offer a SSH shell and a bunch of programming languages including python.
I agree, just noting that there are a lot of people out there who are not that experienced in these kind of operations but are still attracted to using Piwik instead of, say, Google Analytics.
Thank you all. I had managed to figure this out in the mean time.[ul]
[li] The URL is that for the actual piwik installation
[/li][li] idsite is the numeric site id - 1 for the first site
The piwik documentation could have made this a whole lot clearer but I got there in the end.