The connection was lost. Trying to connect again.

Post by Behind The Console

Behind The Console @behindtheconsole@podcasts.social

A Reader for the Archive: Graylog on a Three-Node Proxmox Cluster

Every node already writes home. Somebody has to read the letters.

Since the first part, every node of my three-node Proxmox lab sends its syslog to a log host outside the cluster. Tha

behindtheconsole.com/2026/10/3

#Projekt #ProxmoxCluster #SystemArchitektur #ansible #Graylog #highavailability #homelab #Logging #opensearch #Proxmox #BTC

路 Public

Replies

Behind The Console @behindtheconsole@podcasts.social

@chipiguay yeah the problem is they have one port when they鈥檙e coming. You can take out the SSD and put in a riser card and install a one port or four port card but then you don鈥檛 have to place for the SSD anymore and running only nvme would be not good because I鈥檓 trying to get it running with Nvme and ssd just to be able to build ceph cluster

路 Unlisted

oneuptime @b210beb8f103f785fdccb9e953379174b7999d7b017062fdf179944ff49a758e@mostr.pub
The alert that will matter most here is the absence of logs: a small heartbeat from each node's rsyslog to the log host, and a page when a node has sent nothing for, say, ten minutes. That is the failure mode that quietly turns a working archive into a silent one. The rebuild-time win from 39 to 9 minutes is the kind of thing you only discover because you now have the data to measure it.

路 Public