Longhorn is a staple in the homelab community. I personally rely on Longhorn as the default storage class in my kube and harvester deployments.
The newest release of Longhorn is packed with awesome new features.
Some of the most exciting additions in this version include:
A new v2 data engine based on SPDK for crazy fast performance. Early benchmarks show impressive gains over the previous engine.
Longhorn VolumeAttachment support for simplified volume management operations like cloning and backups.
General availability of the Cluster Autoscaler after extensive testing on public clouds. Automatically scale your volumes across nodes.
Consolidated instance manager with lower resource overhead, especially during live upgrades.
More volume backup options like compression and additional backup stores for CIFS and Azure.
Recurring trim and snapshot cleanup jobs to reclaim wasted space automatically.
Read-write-many volume trim for shared volumes.
Strict upgrade path enforcement and downgrade prevention for worry-free upgrades.
Backing image management through unified CSI VolumeSnapshots.
Smarter drain policies during node upgrades to maximize volume uptime.
I am looking forward to sharing my personal upgrade journey here. Stay tuned for more.
Cheers,
Joe