curl -X POST "https://littlefurnace.com/matomo/index.php?module=API&method=MultiSites.getOne&idSite=2&period=range&date=2024-06-28,2024-08-28&format=JSON" -H "Content-Type: application/x-www-form-urlencoded" --data 'token_auth=???????????????'
curl -X POST "https://littlefurnace.com/matomo/index.php?module=API&method=MultiSites.getAll&period=year&date=yesterday&format=JSON" -H "Content-Type: application/x-www-form-urlencoded" --data 'token_auth=REDACTED'
curl -X POST "https://littlefurnace.com/matomo/index.php?module=API&method=MultiSites.getAll&period=year&date=yesterday&format=JSON" -H "Content-Type: application/x-www-form-urlencoded" --data 'token_auth=???????????????'
​

Cookbook of command line regex examples for sed and grep that worked for me.

1
$ sed -E s/[0-9][A-Za-z]/foooooooo/g  filename.txt

2
New line can be done by piping from sed -> tr
$ echo "123." | sed -E 's/([[:digit:]]*)\./\1|next line/' | tr '|' '\n'

3
Example with "any digit", capture group, and newline
$ echo "123z" | sed -E 's/([[:digit:]]*)z/\1\n next line/'

4
Within  a  bracket  expression, the name of a character class enclosed in "[:" and ":]" stands for the list of all characters belonging to that class.  Standard character class names are:

              alnum   digit   punct
              alpha   graph   space
              blank   lower   upper
              cntrl   print   xdigit

5
How to locate logs
sudo find / -name 'apache2' | grep log
sudo find / -name 'matomo' | grep log | sed /blog/d | sed /sql/d

How to find the fullness of the log
df -h {PATH.TO.LOG}
df -hi {PATH.TO.LOG}

Are you monitoring a website with Matomo? Open a page from your website in a browser, then click browser>>devtools.  The network tab (browser>>devtools>>network) will show Matomo's actual tracking requests.  

A fun extension which also shows this is Chrome Tagbirds. I'm paranoid of browser extensions (because of security), so I only use this on an isolated machine which never runs any of my password secured applications.

I have fomatted these requests by adding linebreaks.

#1
From a single page, send http requests to two Matomo instances at the same time 
As described at https://matomo.org/faq/how-to/how-do-i-send-tracking-requests-to-two-or-more-matomo-servers/ 

The Javascript Tracker Code
<!-- Matomo -->
<script>
var _paq = window._paq = window._paq || [];
/* tracker methods like "setCustomDimension" should be called before "trackPageView" */
_paq.push(['trackPageView']);
_paq.push(['enableLinkTracking']);
(function() {
var u="//littlefurnace.com/matomo/";
_paq.push(['setTrackerUrl', u+'matomo.php']);
_paq.push(['setSiteId', '2']);
var secondaryTracker = 'https://tatll.matomo.cloud/matomo.php';
var secondaryWebsiteId = '6';
// Also send all of the tracking data to the new Matomo server
_paq.push(['addTracker', secondaryTracker, secondaryWebsiteId]);
var d=document, g=d.createElement('script'), s=d.getElementsByTagName('script')[0];
g.async=true; g.src=u+'matomo.js'; s.parentNode.insertBefore(g,s);
})();
</script>
<!-- End Matomo Code -->

Request One
POST https://tatll.matomo.cloud/matomo.php?
action_name=pikL&
idsite=6&
rec=1&
r=357217&
h=11&
m=36&
s=48&
url=https://pikl.us/&
_id=8e2306f5d48d5003&
_idn=0&
send_image=0&
_refts=1714581408&
_ref=https://github.com/atom-box/bpikl&
pv_id=jVb0xW&
fa_pv=1&
fa_fp[0][fa_vid]=Tl9OUd&
fa_fp[0][fa_name]=urlEnterer&
fa_fp[0][fa_fv]=1&
pf_net=0&
pf_srv=266&
pf_tfr=41&
uadata={}&
pdf=1&
qt=0&
realp=0&
wma=0&
fla=0&
java=0&
ag=0&
cookie=1&
res=1920x1080

Request Two
POST https://littlefurnace.com/matomo/matomo.php?
action_name=pikL&
idsite=2&
rec=1&
r=473795&
h=11&
m=36&
s=48&
url=https://pikl.us/&
_id=03690a3479d38e0e&
_idn=0&
send_image=0&
_refts=1714581408&POST https://littlefurnace.com/matomo/matomo.php?
action_name=pikL&
idsite=2&
rec=1&
r=473795&
h=11&
m=36&
s=48&
url=https://pikl.us/&
_id=03690a3479d38e0e&
_idn=0&
send_image=0&
_refts=1714581408&
_ref=https://github.com/atom-box/bpikl&
pv_id=zC90X6&
fa_pv=1&
fa_fp[0][fa_vid]=Tl9OUd&
fa_fp[0][fa_name]=urlEnterer&
fa_fp[0][fa_fv]=1&
pf_net=0&
pf_srv=266&
pf_tfr=41&
uadata={}&
pdf=1&
qt=0&
realp=0&
wma=0&
fla=0&
java=0&
ag=0&
cookie=1&
res=1920x1080POST https://littlefurnace.com/matomo/matomo.php?
action_name=pikL&
idsite=2&
rec=1&
r=473795&
h=11&
m=36&
s=48&
url=https://pikl.us/&
_id=03690a3479d38e0e&
_idn=0&
send_image=0&
_refts=1714581408&
_ref=https://github.com/atom-box/bpikl&
pv_id=zC90X6&
fa_pv=1&
fa_fp[0][fa_vid]=Tl9OUd&
fa_fp[0][fa_name]=urlEnterer&
fa_fp[0][fa_fv]=1&
pf_net=0&
pf_srv=266&
pf_tfr=41&
uadata={}&
pdf=1&
qt=0&
realp=0&
wma=0&
fla=0&
java=0&
ag=0&
cookie=1&
res=1920x1080
_ref=https://github.com/atom-box/bpikl&
pv_id=zC90X6&
fa_pv=1&
fa_fp[0][fa_vid]=Tl9OUd&
fa_fp[0][fa_name]=urlEnterer&
fa_fp[0][fa_fv]=1&
pf_net=0&
pf_srv=266&
pf_tfr=41&
uadata={}&
pdf=1&
qt=0&
realp=0&
wma=0&
fla=0&
java=0&
ag=0&
cookie=1&
res=1920x1080


#2
A heatmaps/screenrecording

PART ONE
This part is optional. 
If you don't need to invalidate reports, skip to Part Two

sudo su www-data -s /bin/bash -c  "php /var/www/littlefurnace.com/matomo/console core:invalidate-report-data"

Matomo parameters are not well-standardized. Here are examples of the syntax for all of the invalidate-report-data parameters I have used this year:
--dates=2024-04-20,today   <---- contrast this to the force-date-range used above
--dates=2024-08-04,2024-08-05
--dates=yesterday 
--periods=day,week
--segment=33
--sites=1,5

Putting it all together, you could run invalidation like this if you wanted to
sudo su www-data -s /bin/bash -c  "php /var/www/littlefurnace.com/matomo/console core:invalidate-report-data --dates=2024-08-04,2024-08-05 --periods=year --segment=33 --sites=1,5"




PART TWO
These are self explanatory hopefully.

If you are in the directory on your machine where the matomo files are, you might be able to just run this, if you are signed in as the right user:
console core:archive

The owner is usually www-data or apache so I usually run it like this:
sudo su www-data -s /bin/bash -c  "php -d memory_limit=3G /var/www/littlefurnace.com/matomo/console core:archive"

Here are examples of the syntax all of the parameters I have used this year:
--concurrent-archivers=1 <---- often won't work; minimum of 2, on many systems
--concurrent-archivers=5
--concurrent-requests-per-website=1
--concurrent-requests-per-website=3
--disable-scheduled-tasks
--force-date-last-n=220
--force-date-range="2023-01-01",today
--force-date-range=2023-09-01
--force-date-range=2023-09-01,2023-09-01
--force-date-range=yesterday,today
--force-idsegments=12
--force-idsites=2,3,11
--force-sites=2
--skip-all-segments
--vvv

I have never run all at the same time, but you could put most of that together and run the following. (I left out the two segment options which would contradict each other if run at the same time, although in that case, the last parameter entered would possibly be the boss ... later parameters take priority over earlier ones usually.)

sudo su www-data -s /bin/bash -c  "php -d memory_limit=6G /var/www/littlefurnace.com/matomo/console core:archive --concurrent-archivers=2 --concurrent-requests-per-website=3 --disable-scheduled-tasks --force-date-last-n=220 --force-idsites=2,3,11 --skip-all-segments --vvv"

​
PART THREE
My personal Matomo is running on a underresourced machine: one CPU, 4G of RAM. So I get to see all of the error messages, as follows.
sudo su www-data -s /bin/bash -c  "php ./console core:archive --force-idsites=1  --force-date-last-n=20"
If you run core:invalidate-report-data and then run the above command, twice in rapid succession, nothing much is done the second time. Each step took 3 seconds. 

sudo su www-data -s /bin/bash -c  "php ./console core:archive --force-idsites=2  --force-date-last-n=1"
Crashed ("unserialization error" and " Warning - shell_exec(): Unable to execute 'ps x 2>/dev/null" and "Unable to execute 'ps aux'"). Half of these mentioned a segment.
sudo su www-data -s /bin/bash -c  "php -d memory_limit=9G  /var/www/littlefurnace.com/matomo/console  core:archive   --concurrent-requests-per-website=1 --concurrent-archivers=3 --skip-all-segments --force-idsites=2"
Succeeded.
Ran all of the Custom Reports, back til the beginning of time for the new CR I authored recently.
sudo su www-data -s /bin/bash -c  "php -d memory_limit=9G  /var/www/littlefurnace.com/matomo/console  core:archive   --concurrent-requests-per-website=1 --concurrent-archivers=3  --force-idsites=2"
Succeeded.
I think the segments worked because I had already run all the basic reports.

If you are a hater of cookies, Matomo is able to run with no cookies. There is a setting in the user interface: cookies are easy to disable.*

But cookies are enabled by default. Why? The Matomo Visitor ID cookie is a first party cookie! Look it up: that's a respectful, ethical type of cookie.  And the data it collects goes to your website, only.  So if you are trying to keep a lot of secrets about what goes on in your website, Matomo is the way to do it. 

Matomo explicitly promises that they have no interest in your data. See these:
https://matomo.org/100-data-ownership/
https://matomo.org/matomo-cloud-privacy-policy/ (Matomo Cloud only)

How to see the Visitor ID, and confirm that Matomo sets a cookie during the visit:

1. Go to your website in a browser.
2. Open devtools >> Application/Storage/Cookies
3. pk_id is the visitor ID. Its value is a 16 char string.
4. Go to your Matomo dashboard.
5. You can see this 16 char string in Visitors >> Visits Log (see the screenshot above)

*What Matomo loses when you disable the Matomo cookie, is it cannot detect a returning visitor. This messes up some of the more subtle things you would want to know about your website. Returning visitors are a pretty big deal to understand.

Here is something you can try in the console of any browser if you are visiting a page which uses the Matomo Tracker JS.

1. Open:
   Browser >> Dev Tools >> Console
2. Enter these 4 commands one at a time in this order, hitting <ENTER> after each one
3. _paq.push(["setUserId", anExampleVariable]); <ENTER>
4. let anExampleVariable = "The Man in the Moon"; <ENTER>
5. _paq.push(["setUserId", anExampleVariable]); <ENTER>
6. _paq.push([ function() {console.log(this.getUserId());}]); <ENTER>

Step 6 allows you to test whether Matomo stored the name of the User ID.

Then let's do this using your website where you have stored the name of your website visitor. We are going to take the visitor's name and store it as their Matomo UserID.

Let's assume your website allows the user to sign in, with a password, and the username is stored in a JS variable in the browser called first_and_last.

In a browser which has your user's web page load your website. Sign in as a username in your website:

1. Browser >> Dev Tools >> Console
2. Enter these 3 commands one at a time in this order, hitting <ENTER> after each one
3. first_and_last; <ENTER> just to see if your name is already in the page, nothing to do with Matomo yet
4. _paq.push(["setUserId", first_and_last]); <ENTER>
5. _paq.push([ function() {console.log(this.getUserId());}]); <ENTER>

Setting the UserID in Matomo has benefits:

• You can see the UserID immediately in Dashboard >> Visits >> Visits Log
  
• You can see a special report on signed in users at Dashboard >> Visitors >> User IDs
• You can detect each history of each visitor on your site, including the monetary value of their conversions.  (You have to set up the conversions manually - see links below. Once you set these up, you can start measuring the rich data of how often and how much are the conversions worth, over time, for this visitor.)
  

Matomo uses three ways to string together repeated visits over time for a visitor. The weakest but most legal (most private) is the scrambled fingerprint number. This is remembered only for today (actually less than today: for privacy, this ID is deleted within 30 minutes of the session end.) You can make this time longer than 30 minutes but it puts a high demand on your CPU if you try to set it for several days duration. Still, this might work for a low traffic website.   It will not work for Matomo for Wordpress: those deployments have small resources in the hardware: Matomo has to share CPU and SQL database with the WP resources. Not good.

Stronger than this is detection by first-party cookies.  By default, Matomo will do this. If you want to be super private with your visitors data you should disable this.

Strongest of all is detection by User ID. UserID is what we experimented with at the top of this article.

Read more at the official Matomo Docs:
How does Piwik detect unique and returning visitors? (with User ID or Visitor ID from tracking cookie)
https://matomo.org/guide/reports/user-ids/
https://matomo.org/guide/reports/goals-and-conversions/

Goal:
Pedagogically poke around: get to know how Matomo attaches Javascript to your website.

Method:
We will just "mess around".

Warmup:
Open a web page which is running Matomo Javascript Tracker.

Feel free to open your own website. Or you could load your browser with a public website that uses Matomo's Javascript tracker.  At the time of this writing, this includes, westerndigital.com, bundesliga.de, oebb.at, msufcu.org, telummedia.com, bcferries.com, and 1.6% of the internet.**

Lesson 1:
Open any website which has javascript from Matomo already (see above).
Open the browser dev tools >> console (I am using Firefox devtools, so adapt if you're on Chrome/Safari/Brave/Arc browser)
In Console type _paq
You are now seeing the Piwik object. Objects are JS data structures, similar to arrays but they can nest. And instead of just storing strings and numbers, they can store executable code. Notice that most of the _paq object is functions.  This _paq is named for Piwik still, from when that was Matomo's old name. (There is a history link at the base of Matomo's home page if you are curious.)

Do not try to store values in _paq. It is intended to be a place to run commands, as in _paq.push();

Lesson 2:
You can see all available methods by entering this in the browser console... window.Matomo.getAsyncTrackers()[0]

Optional Reading: https://developer.matomo.org/guides/tracking-javascript-guide
Optional Reading: https://developer.matomo.org/api-reference/tracking-javascript

**Some web pages use Matomo without JS. (Turns out that Javascript is just one of the four ways to track visitors at a website, but JS is the most common, by far.) And in some cases there will be no Matomo code unless you click Yes on a consent banner.

How to 99% of everything your Matomo dashboard knows is stored and retrieved from its MySQL database. Here are a half dozen useful queries:

LOOKING FOR A COLUMN BY NAME, LIMIT IT TO ONLY ONE DB
SELECT TABLE_NAME, COLUMN_NAME name FROM information_schema.columns WHERE COLUMN_NAME LIKE "%key%" AND table_schema = 'XYZfurnace_mat_jan2022';

DELETE THE CONTENTS BUT NOT THE TABLE
In SQL, to delete the rows:
TRUNCATE TABLE table_name;

INSPECT THE RAW DATA
If you want to directly check for the raw data in the SQL database you could also run a query for the date, similar to
SELECT * FROM matomo_log_visit WHERE visit_first_action_time > '2023-10-21' AND visit_first_action_time < '2023-10-22';

INSPECT RAW DATA FROM THE DASHBOARD ITSELF.
1) Dashboard >> Visitors >> Visits log    (Tableau de bord >> Visiteurs >> Journal des visites)
OR
2) Dashboard >> ⚙️  >> Platform >> API  (Tableau de bord >> ⚙️  >> Plate-forme >> API)

SEE EVERY COLUMN NAME IN THE ENTIRE DATABASE
This is a great flyby of the entire data table. You can also get the lay of the land by simpling searching online for matomo data schema. SQL feels more exciting because you're seeing it in situ.
SELECT TABLE_NAME, COLUMN_NAME name FROM information_schema.columns ORDER BY COLUMN_NAME;

SEE EVERY COLUMN NAME IN THE ENTIRE DATABASE SIMILAR TO ID VISIT
This is useful if you think you know a variable name from a cookie or from the data layer or something and you wonder where is it stored in the database.
SELECT TABLE_NAME, COLUMN_NAME name FROM information_schema.columns WHERE COLUMN_NAME LIKE '%id%visit%';

MORE QUERIES
For lots of great SQL query recipes, see the official Matomo SQL cookbook:
https://matomo.org/faq/how-to/how-do-i-write-sql-queries-to-select-visitors-list-of-pageviews-searches-events-in-the-matomo-database/

How to see the scheduled tasks in Matomo.
Use SQLto see the scheduled tasks in Matomo (these run at the end of any core:archive run or can be run manually with core:run-scheduled-tasks.

1. In the db, SELECT * FROM matomo_option WHERE OPTION_NAME LIKE "%sched%" ;
2. Break the lines on semicolon.

foooo
fooooo
​barrrrrrrrrrr