Matomo#

Matomo is the web analytics platform.

Available at: https://analytics.openfoodfacts.org/

You must have a user account to access it (hopefully !). Ask for an admin to create you an account if you need it (Beware, there are personal information in the sense of GDPR like ip addresses). Ask for it to contact email.

Main running services#

nginx is used as a HTTP frontend
php7.3-fpm run the matomo software
mariadb is the main database for matomo, some configurations are made for performance (see /etc/mysql/mariadb.conf.d/90-off-configs.cnf )
redis is used to fast track matomo requests (see Matomo setup for performance, below)
different systemd timer takes care of putting tracking from redis to mariadb and to consolidate archive reports (see Matomo setup for performance, below)
prometheus exporters are installed for nginx and mysql (see Prometheus exporters, below)

Most of those systemd services should email on failures.

Important configuration files are linked to this git repository (see using git, below)

Alerts#

You can setup alerts in matomo. In personal alerts.

We have setup alerts for main website and app if there are 0 visits.

Important: temporary patch#

Because of Queue id should try to be more random #231 and Set a random visitor ID at each requests instead of using 0000000… for anonymous analytics

until enough mobile app are deployed, I manually patched plugins/QueuedTracking/Queue/Manager.php:

    protected function getQueueIdForVisitor($visitorId)
    {
        # 2024-03-08 patch from ALEX
        if ($visitorId === '0000000000000000') {
            $visitorId = chr(rand(ord('a'), ord('z')));
        }
        $visitorId = strtolower(substr($visitorId, 0, 1));
      ...

Site setup#

goto manage / websites and add a website

To be GDPR compliant (and user friendly) ¹:

in your Matomo Tag, you can check the option « Disable cookies » which will disable all first party tracking cookies for Matomo. ²
To ensure that you do not store the visitor IP, which is Personally Identifiable Information (PII), please go to Administration > Privacy > Anonimyze data, to enable IP anonymization, and check you have 2 bytes or 3 bytes masked from the IP address. ³

In productopener#

We use the $google_analytics variable in config to add the javascript snippet for Matomo.

Matomo install#

using git#

We use git to track some important configurations files in this repository.

The repository is checked out in /opt/openfoodfacts-infrastructure and specific /etc configurations files are symlinked there (see confs/matomo).

Matomo setup for performance#

We setup matomo for performance (our websites requires it) with two main points:

it does not process archives on incoming requests but instead on a systemd timer (see confs/matomo/systemd/matomo-archive.{service,timer}). See also official doc.

on incoming update request (on a tracked website being visited), it does not immediately updates the database but goes in redis instead, we configured it with 8 queues. then are systemd timers jobs to process redis entries every minute (see confs/matomo/systemd/matomo-tracking@.{timer,service}). And we have an instance 0, 1, 2,… 7, to process each queue. See also official doc

MariaDB has been tuned a bit toward performance (using more memory) see /etc/mysql/mariadb.conf.d/90-off-configs.cnf (linked to this repository confs/matomo/mysql/mariadb.conf.d/90-off-configs.cnf)
- we also tried to avoid "2006 MySQL server has gone away" following https://matomo.org/faq/troubleshooting/faq_183/

Archiving is splet in more than one systemctl timer / service, by using matomo-archive@<name>.timer/service which correspond to different options it /etc/matomo/archive-<name>.env, so that:
- there is one archiver specialized for the openfoodfacts.org website
- one specialized for the mobile app
- one for all the other websites
( all this thanks to --skip-idsites/--force-idsites options)

Both tracking and archiving logs to /var/log/matomo.

Prometheus exporters#

Nginx prometheus exporter is installed and needs the stub_status site, which exposes nginx metrics.

Mysql server prometheus exporter is installed (for MariaDB), its /etc/default configuration file sets the connection string. The corresponding mysql user had to be manually created (instrutions in the config file).

Updating the Matomo version#

Just use the web administration to update the software.

Setup robots.txt to avoid search engine indexing#

Setup /var/www/html/matomo/robots.txt as is:

User-agent: *
Disallow: /

How to#

How to investigate 0 visit problems#

You first have to see if the request made by a browser to matomo returns a 204 No Content response (use the network panel of developer tools).

Also verify that the site_id is correct.
If it's ok but you don't see the visit, you may look at real time visits of your website. If there is no data it may means the tracking jobs are not running correctly.

You can log on the server and use mysql matomo_db to look at the database, in particular matomo_log_visit table contains visit logs.

Beware that it's a huge table, try to limit your query with something like WHERE idsite=5 AND visit_last_action_time > '2024-01-16'

You can also see section below How to monitor tracking jobs.
incomplete graph might also come from archive processing. Sometimes, in this case, scaling down to days might show you right results, whereas you get incomplete result at the week or month level.

See How to monitor archive jobs.

How to monitor tracking jobs#

When a request arrive, it is logged into redis. We then have matomo-tracking systemd services to process the redis queue.

To see status:

# see jobs status
systemctl status matomo-tracking@{1..3}.service
# see timers status
systemctl status matomo-tracking@{1..3}.timer

Matomo also offers some commands to follow the status of the tracking jobs:

cd /var/www/html/matomo
./console queuedtracking:monitor

# you might also look at lock status
./console queuedtracking:lock-status

use --help for more options.

Logs are in /var/log/matomo

How to monitor archive jobs#