This message was deleted.

• Do you only have one storage disk per node? I would suggest 3 disks minimum. • Check networking between nodes, ssh into machines and make sure they communicate properly. • Login to the longhorn backend and see what is going on there. You can do this under support on bottom left.

Thanks for the recommendations. • I cannot add additional disks to cluster. This is a small form factor PC test lab, that I used to learn prior to doing more in a production environment. The current testing is looking into replacing VMWare with Harvester. in my Prod VMWare environment we are using vSAN so that lead me to trying harvester because of the native longhorn integration. • I can ssh between all of the nodes. • Everything in the longhorn dashboard is showing as schedulable and healthy I am going to deploy some more test VMs and see if anything pops up in Longhorn

First thing I do is tune ethernet adapters using ethtool on hosts and make sure any bridge interfaces have mtu 9000, max ring and buffers, and update things like txqueuelen. After that I setup networks and vlans including one for storage. Next I update any longhorn settings available in the harvester UI, like storage network. This is exactly what you would do with vsan. Then I go into longhorn directly and update severl settings. In your case with only one disk per node this is likely a good place to solve your issue. Some settings to take a look at are node and zone soft affinity, data locality, replica auto balance, etc.

Thanks for all of the advice. I will work through it methodically over the next couple of days.