Pulse not showing more than 2 docker hosts even after 3 or 4 hosts are added

[Radixb]

Describe the bug
I have 4 Docker hosts. Added the first 2 with unique API tokens and they show up without any issues. After additional hosts beyond those two are added, docker tab still only displays 2 hosts and when the agents from hosts 3 and 4 report, they briefly display. This appears to eventually crash Pulse docker monitoring. ** this possibly is related to the enhancement request just added although I don't recall cloning my vm**

To Reproduce
add a 3rd or 4th docker host

Expected behavior
I would expect host count to be correct, and all hosts to be displayed.

Environment:
Pulse | Version: 4.24.0 - Docker

Additional context
Add any other context, screenshots, or logs here.

💡 Tip: If you're experiencing connection issues, API errors, or missing data, you can attach diagnostics from Settings → Diagnostics tab → Export for GitHub (sanitized version)

[Radixb]

This was caused by duplicate engine-ids in my docker hosts. I was able to fix this by first deleting the machine-id file from both /etc/machine-id and /var/lib/dbus/machine-id and rebooting, then rebuilt it using command 'systemd-machine-id-setup'
then renaming /var/lib/engine-id to engine-id.old and issued command 'systemctl restart docker to reissue the engine-id file.

[PavelNL]

@rcourtman: I'm getting even worse situation in today's fresh install of Pulse:

• there are two PVE Docker VM's with agents installed via commandline with two different Tokens.
• they are switching in output in Docker tab in Pulse, but never shown both at the same time.

Do I need to create a new issue or just wait a bit till the "Add collision-safe Docker host identifiers" patch is released?

PS: Indeed, one VM was a clone of another, but I've updated machine-id as in @Radixb post above.

[Radixb]

@rcourtman: I'm getting even worse situation in today's fresh install of Pulse:

• there are two PVE Docker VM's with agents installed via commandline with two different Tokens.
• they are switching in output in Docker tab in Pulse, but never shown both at the same time.

Do I need to create a new issue or just wait a bit till the "Add collision-safe Docker host identifiers" patch is released?

PS: Indeed, one VM was a clone of another, but I've updated machine-id as in @Radixb post above.

A couple things to try--run the 'docker info' command and check that the id has indeed changed to be unique. I had to delete both the machine.id and engine.id files. I also rebooted the vm, and I removed the Pulse tokens and rebuilt them. Not sure the latter is necessary, but easy enough to try.

[adirburke]

I had the same issue, I cloned a VM with docker pre-installed so all the machines had the same docker machine id, even thought they had different host machine id's. I added the env - PULSE_AGENT_ID=${PULSE_AGENT_ID} and populated the agent id from /etc/machine-id

[PavelNL]

A couple things to try--run the 'docker info' command and check that the id has indeed changed to be unique. I had to delete both the machine.id and engine.id files. I also rebooted the vm, and I removed the Pulse tokens and rebuilt them. Not sure the latter is necessary, but easy enough to try.

You are right, Docker ID is the same for both hosts:

docker info|grep ID

ID: 144*

That was enough:

mv /var/lib/docker/engine-id /var/lib/docker/engine-id.old ; systemctl stop docker ; sleep 10 ; systemctl start docker ; docker info|grep ID

ID: 19b*

And, indeed, the issue is resolved. Thank you!

[PavelNL]

I had the same issue, I cloned a VM with docker pre-installed so all the machines had the same docker machine id, even thought they had different host machine id's. I added the env - PULSE_AGENT_ID=${PULSE_AGENT_ID} and populated the agent id from /etc/machine-id

Do I understand it right, that you are running the Pulse Agent in a container instead of systemd ?
Is it really much better to do it this way vs. the official way as represented in the WebUI?

[adirburke]

My pulse was saying the version i installed from the WebUI way was out of date and to upgrade so I tried the docker version, It was also very easy for me to deploy, I wrote a bash script take takes a csv of the VM's IP and their key and created the docker compose file and ran it. Was very quick for me to do it.

Can supply script if you want

[PavelNL]

My pulse was saying the version i installed from the WebUI way was out of date and to upgrade so I tried the docker version, It was also very easy for me to deploy, I wrote a bash script take takes a csv of the VM's IP and their key and created the docker compose file and ran it. Was very quick for me to do it.

Can supply script if you want

Thank you. Already made the Gitea repo and user DockerHub private repo to get the image delivered to Portainer.