How-to monitor all site processes

Owned by Matthew Hierholzer
Last updated: about 2 hours ago

Monitoring the Checkmk site's processes, like cmc and helpers, is useful for debugging performance issues, spotting memory leaks, and responding quickly to outages or missed notifications.

LAST TESTED ON CHECKMK 2.3.0P1


Table of Contents

Step-by-step guide for monitoring all Site processes


To configure all relevant site processes to be monitored, follow these steps:

  1. Create a Process discovery rule:

    Image showing the options on process discovery

    • The Process name is:
      "%u/%1%2"

    • The regular expression is:
      (?:(?:python3 |\/usr\/sbin\/apache2 -f )*\/omd\/sites\/\w+\/.*?(checkhelper|icmpreceiver|apache|--create-rrd|--handle-alerts|(?:mk)*notifyd(?:d)*|--notify|--checker|--real-time-checks|fetcher|mkeventd|rrdcached|redis|cmc |icmpsender)|(dcd|gunicorn|liveproxyd))

      Make sure to enable the "Grab user from found processes" option.

      The rule should match your Checkmk server hosts.

    .

  2. After saving and activating the changes, do a service discovery on your Checkmk server hosts. There, you will see a bunch of services to be added:

    Image showing Checkmk servies

    Click on “Accept all”  to add them to the monitoring.

    After activating the changes, your site's processes will be monitored in a basic way, already.


    .

  3. Some of these processes should only run once, so let's add a "State and count of processes" rule for them:

    Image showing the count and processes screen

Alternative regular expressions for monitoring particular Site processes

We recommend monitoring all site processes, but if you want to pick particular sets of processes, you can use these regular expressions:

ComponentRegular Expression
Active check helpers
\/omd\/sites\/\w+\/lib\/cmc\/(checkhelper)
Agent receiver
(gunicorn)
Alert handler
python3 \/omd\/sites\/\w+\/bin\/cmk .*?(--handle-alerts)
Checkers
python3 \/omd\/sites\/\w+\/bin\/cmk .*?(--checker)
DCD(dcd)
Event console
python3 \/omd\/sites\/\w+\/bin\/(mkeventd)
Fetchers
python3 \/omd\/sites\/\w+\/bin\/(fetcher)
ICMP receiver
\/omd\/sites\/\w+\/lib\/cmc\/(icmpreceiver)
ICMP sender
\/omd\/sites\/\w+\/lib\/cmc\/(icmpsender)
Liveproxy daemon(liveproxyd)
Microcore
\/omd\/sites\/\w+\/bin\/(cmc)
Notification helpers
python3 \/omd\/sites\/\w+\/bin\/cmk .*?(--notify)
Notification spooler
python3 \/omd\/sites\/\w+\/bin\/(mknotifyd)

RRD cache daemon

\/omd\/sites\/\w+\/bin\/(rrdcached)
RRD helpers
python3 \/omd\/sites\/\w+\/bin\/cmk .*?(--create-rrd)
Real-time helper
python3 \/omd\/sites\/\w+\/bin\/cmk .*?(--real-time-checks)
Redis DB
\/omd\/sites\/\w+\/bin\/(redis-server)
Site Apache\/usr\/sbin\/apache2 -f \/omd\/sites\/\w+\/etc\/(apache)\/apache.conf


Each of these expressions contain one group, so you can use "%u/%1" as Process name in your Process discovery rule.