This message was deleted.

Yes, some people are using it in production. But you need to test it for a while to understand what you can and cannot do, so that you won't break the cluster 🙂

I used it in the past in fact, trying to catch with it again. But the biggest bottleneck is longhorn sometimes as it seems to rebuild volumes each times so maybe the new version fix that. I don't have a poor cluster ( 10GBit links and SSD disks on recent dell hardware )

Longhorn is slow for us too. Not running at full potential/speed of the SSDs. We have the cluster nodes on 100 Gbps network. Not sure yet what is causing the performance issues.

how does longhorn sync data between replica ? Because on single replica it works ok.

Yes, We are running a 20 node harvester cluster on with 100Gbps nics, 1TB of memory, 5 nvm disks, and 128core per node we would love to offer our input.