// Blog · tag

#high availability

Every post tagged high availability.

cluster

Tuning Corosync Link Priority and Token Timeouts for Proxmox VE Cluster Resilience

Proxmox VE corosync redundancy only works if knet actually prefers the right link. Here's how to set link priority correctly, read corosync-cfgtool output, and avoid tuning the token timeout blind.

cluster

Setting Up a Corosync QDevice for Two-Node Proxmox VE Cluster Quorum

A two-node Proxmox VE cluster loses quorum the moment one node reboots. Here's why the vote math breaks down and how to fix it permanently with an external corosync QDevice as a tie-breaker.

cluster

What Is a Proxmox Cluster? (And Do You Actually Need One?)

Clustering sounds like something only data centers need, but it's really just several Proxmox nodes agreeing to share one view. Here's what Corosync and quorum actually mean, and when you genuinely need a cluster versus when you don't.

high-availability

Disarming HA in Proxmox VE 9.2: Safe Cluster-Wide Maintenance Without Fencing

Proxmox VE 9.2 adds cluster-wide disarm-ha and arm-ha commands so you can do network maintenance on an HA cluster without triggering fencing. Here's how the feature works and how to use it safely.

high-availability

Configuring Redundant Corosync Links for Proxmox VE Cluster Resilience

A single corosync link is a hidden single point of failure in every Proxmox VE cluster. Here's how to diagnose link instability and configure a second, truly independent corosync link using knet.

storage

Diagnosing and Fixing Failed Proxmox VE Storage Replication Jobs

ZFS-based replication keeps Proxmox VE HA working without a SAN, but jobs fail quietly from broken snapshot chains, renamed storage IDs, and full target pools. Here's how to read pvesr's errors, find the real root cause, and trust the GUI again.

cluster

Troubleshooting Corosync Cluster Communication Failures in Proxmox VE

A practical field guide to diagnosing and fixing corosync communication failures in Proxmox VE clusters, covering symptoms, root causes, diagnostic commands, redundant link configuration, and best practices to prevent HA fencing incidents.

cluster

Preventing Split-Brain in Two-Node Proxmox VE Clusters with a Corosync QDevice

How to deploy a Corosync QDevice so a two-node Proxmox VE cluster keeps quorum and High Availability working through a single node failure.