Was looking at the client connections on an freshly rebooted Isilon cluster and noticed that they were not balanced..not even close.
All connections are SMB so SmartConnect Advanced does not come into play. SmartConnect basic is configured as Round Robin and all nodes have the same # of IP's assigned.
"Always use the fully qualified domain name (FQDN) of a SmartConnect zone when accessing the cluster. If you attempt to use the short name, Windows hosts will attempt to use the NetBios name service (NBNS) to resolve the connection. Because NBNS uses broadcast pings on the network to determine what IP a host is located at, the Windows client will connect to the first node to respond, which might result in client connections not being evenly distributed across the cluster. Additionally, by using the NBNS services, you do not utilize Kerberos for authentication and authorization, and are required to use NTLM (NT LAN Manager) based services, which can lead to permission denied errors."
I opted for disabling NetBios via our DHCP server (you can also do it with a regedit via GPO), rebooted the client machines, and all is well.
Couple notes:
- Netbios is still needed to initially join a machine to a domain so keep this in mind for your imaging/deployment process. Once joined it can be disabled.
- Netbios/WINS are required for certain legacy applications. Have not run into any issues yet but something to keep in mind especially for larger deployments.
- Netbios without WINS is incredibly chatty. If you can get away with not having it on your network you probably should.
- If you do decide to keep NetBios/WINS on client machines, utilize drive mapping and make sure that mapped drives use the FQDN (should map to \\isilon.mycompany.local not \\isilon)
- If you do keep NetBios, check your pipeline/scripts for UNC references to your isilon cluster...there might be a few in there that are lacking a FQDN.
--