[Bug] Syslog view per Host

[buzzzo]

---

Describe the bug

On $proxmox_node->resources->syslog i would expect to displat ONLY the syslog's entries of this host.
Instead ALL host's log entries are displayed.

Steps to Reproduce

1. Select a proxmox node/cluster
2. Go to Resources, then syslog
3. You will see ALL syslogs entries for ALL hosts managed by pegaprox

Expected behavior

ONLY Selected Host/Cluster's syslog entries should be displayed.

Environment

• PegaProx Version: 0.9.9.3
• Installation Method: manual/script
• OS: debian 13
• Browser: chrome
• Behind Reverse Proxy: yes, nginx
• UI: corporate

Screenshots

Checklist

• [Yes ] I have searched existing issues to make sure this is not a duplicate
• [ Yes] I am using the latest version of PegaProx

[mkellermann97]

Confirmed. The "Resources → Syslog" view in the corporate UI is fetching from the integrated PegaProx syslog server (/api/syslog/events), which collects entries from every host pushing to the centralised endpoint — so when you scope the page to one node it should be auto-filtering on hostname=<node> and isn't.

The backend already supports it: /api/syslog/events accepts hostname, source_ip, severity, protocol, facility, search etc. as query params. Per-host scoping just needs the frontend to wire hostname into the request when the user has a node selected.

Easy frontend-only fix. Adding to the queue. Workaround until it lands: in the Filter row above the syslog table, type the node hostname into the host-filter box manually — the backend pipes it through correctly.

Tagging bug and corporate-ui.

https://opencollective.com/pegaprox

— Marcus

[gyptazy]

Hey,

It should be in the scope of the cluster, representing the content of all nodes, just like the snapshots. That was at least the idea when adding this feature. As mentioned, you can still filter for a specific node name. But maybe pre-filtered to the current selected host could be added.

[buzzzo]

If you see my screenshot you would note that the node selected show syslog entries for hosts that aren't member of the current selected one.

And to tell the truth the selected node does not have syslog forwarding at all meaning that it hasn't any logs.

[gyptazy]

If you see my screenshot you would note that the node selected show syslog entries for hosts that aren't member of the current selected one.

And to tell the truth the selected node does not have syslog forwarding at all meaning that it hasn't any logs.

Fair enough, it was more about sharing the ideas that it should be treated in the cluster scope. At least that was the initial idea when I added this feature.

[gyptazy]

Hey @buzzzo,

what do you think about such a solution, where administrator can change the general behavior of this. I'm aware, that already several people using this as a centralized logging solution to get all logs from all their systems there. By a strict change, this would become unusable for them.

As a result, I think we can simply add a bool to the setting within a new chapter where people can set a google to see all logs or simply to scope it to the nodes of the selected cluster. This would look like this and is wired to the front- & backend:

When the scoping is not activated, people can still use it as a centralized solution to also get the logs from other workloads like:

Happy to hear some feedback for this and hope this fits all needs.

Cheers,
gyptazy

[buzzzo]

This makes perfectly sense.
And you could add a button on the syslog viewer "show all" which reset all filters in order to show all hosts syslog entries.