Discussion thread for official Zabbix Template Docker

[AlexL]

This thread is designed to provide grounds for discussion of the official Zabbix Template for Docker.
The template and details of the template is available in GIT repository: https://git.zabbix.com/projects/ZBX/...tes/app/docker

Zabbix is always looking for ways to improve our services and to make our users happier.
We pride ourselves on doing our best each and every day, but we know that there is always something more to learn.
We would like to hear back from you to know what have you liked and what would you improve in the template.

[eronisko]

Thanks for this template.

We'd like to use this to monitor our Swarm services. Is there anyway we could extend this template to include (additional?) calls to the Docker API, e.g. https://docs.docker.com/engine/api/v...on/ServiceList ?

[andrii.katsubo]

Nice template! Is it possible to use it with Zabbix Agent 2 on Windows Server 2019?

[Shaman0S]

Thanks for the thing, allowed me to move out from self-baked solution.

The only thing I am missing is monitoring amount of threads/processes in a container.

When you execute 'docker stats' you can see it in 'PIDS' column. Also this metric item exposed via Docker API since v1.23.

It helped me and my colleagues to identify multiple thread-leaking apps in production. So I love to see it in one of upcoming releases and created feature request ZBXNEXT-8354

Anyone who's missing that, please vote.

[DrMoe]

Hi thanks for the template but it unfortunately doesnt work as expected with Docker swarm.

It discovers the containers but since the naming in docker swarm is randomly generated.
For example if you restart a set of containers it correctly detects that the container is down but cannot detect that the container is up because the Name of the container is not the same.

Example:
curl -s --unix-socket /var/run/docker.sock [URL]http://dummy/containers/json[/URL] | jq '.[].Names'

Example Output BEFORE restart:
[
"/servicefrontend.1.os4eb7466kac5yr55tr9oiz9k"
]
[
"/servicefrontend.2.f6aszyhb0du0xu8fehz520v9f"
]​

Example: Output AFTER docker container restart:
[
"/servicefrontend.1.x1jhvjjf79r2e3nalz5kr0f45"
]
[
"/servicefrontend.2.5jrn722eh7dhj7n0h2bd0nr7k"
]​​

Can this be fixed for docker swarm? it just needs to cut off the randomly generated part.

So the discovery should only discover:

servicefrontend.1
servicefrontend.2

[gelowe]

I am having trouble getting Docker Template to work with Podman
I get the error : dial unix /var/run/docker.sock: connect: permission denied.
How do I allow Zabbix to read the socket file?

In podman the docker sock file is
ls -lht /var/run/docker.sock
lrwxrwxrwx 1 root root 23 Nov 29 2022 /var/run/docker.sock -> /run/podman/podman.sock
ls -lht /run/podman/podman.sock
srw-rw---- 1 root root 0 Nov 29 2022 /run/podman/podman.sock

[DrMoe]

I am having trouble getting Docker Template to work with Podman
I get the error : dial unix /var/run/docker.sock: connect: permission denied.
How do I allow Zabbix to read the socket file?

In podman the docker sock file is
ls -lht /var/run/docker.sock
lrwxrwxrwx 1 root root 23 Nov 29 2022 /var/run/docker.sock -> /run/podman/podman.sock
ls -lht /run/podman/podman.sock
srw-rw---- 1 root root 0 Nov 29 2022 /run/podman/podman.sock

Try
usermod -a -G docker zabbix
service zabbix-agent2 restart

[gelowe]

There is no docker group on a server with Podman , so this won't work. I don't think creating the group will help either as podman does not recognize the docker group.

I did find a way, by adding the zabbix group the the podman socket
I copied /usr/lib/systemd/system/podman.socket to /etc/systemd/system/podman.socket
I added line SocketGroup=zabbix


[Unit]
Description=Podman API Socket
Documentation=manodman-system-service(1)

[Socket]
ListenStream=%t/podman/podman.sock
SocketMode=0660
SocketGroup=zabbix

[Install]
WantedBy=sockets.target

then i reloaded systemd and restarted the socket
systemctl daemon-reload
systemctl restart podman.socket

[markfree]

This works for root, but if you try it with a rootless user, it fails with a chown error.

[flajzi]

I have interesting behavior of Docker plugin.

When I try ping Docker service locally by Zabbix Agent 2, I receive "1":

zabbix_agent2 -t docker.ping -v
...
2023/11/19 14:23:04.000723 [2] processing update request (1 requests)
2023/11/19 14:23:04.000774 [2] registering new client
2023/11/19 14:23:04.000781 Calling C function "new_global_regexp()"
2023/11/19 14:23:04.000797 [2] adding new request for key: 'docker.ping'
2023/11/19 14:23:04.000806 [2] created direct exporter task for plugin 'Docker' itemid:0 key 'docker.ping'
2023/11/19 14:23:04.000823 [2] created configurator task for plugin Docker
2023/11/19 14:23:04.000844 plugin Docker: executing configurator task
2023/11/19 14:23:04.000904 executing direct exporter task for key 'docker.ping'
2023/11/19 14:23:04.001422 executed direct exporter task for key 'docker.ping'
docker.ping                                   [s|1]
2023/11/19 14:23:04.001459 manager has been stopped​

but when I do the same from remote Zabbix server by calling docker.ping to SAME Zabbix agent2 I get "0" and in log I see this:

2023/11/19 14:18:06.846904 received passive check request: 'docker.ping' from '172.16.238.2'
2023/11/19 14:18:06.846955 [1] processing update request (1 requests)
2023/11/19 14:18:06.846967 [1] adding new request for key: 'docker.ping'
2023/11/19 14:18:06.846976 [1] created direct exporter task for plugin 'Docker' itemid:0 key 'docker.ping'
2023/11/19 14:18:06.847006 executing direct exporter task for key 'docker.ping'
2023/11/19 14:18:06.847140 executed direct exporter task for key 'docker.ping'
2023/11/19 14:18:06.847174 sending passive check response: '0' to '172.16.238.2'
​

Can anyone explain why is this happening?

[autumnwalker]

Hello! Thank you for this Template!

Is there a way to have this template monitor for updates available to individual containers?

[IgorPMAAA]

Hi, sorry I am new to Zabbix.

I have configured root-less docker in SLES15SP6 following this documentation.
Is there anything to adjust or another template to make it work with root-less docker?

Thanks

[gaoquan_yang]

hello.the container memory usage does not match docker stats command. do you have any suggestion ?thanks

[hadesinua]

Is there an active version available to monitor active only agents?