DHCP Issues With Aruba Switch

I’ve been slowly building up my home network over the past few months thanks to all the information and sharing from this community.

At this point my network looks like:
Telco => pfSense (running DHCP) => Aruba S1500
Connected to the S1500 I have an Unraid server, Plex server running Ubuntu, 3 IAP-207s and various clients (desktop, media players, etc.).

The Aruba S1500 was factory reset when I received and I did some basic configuration to setup the management interface. Other than VLAN1 there are no additional VLANs or other segmentation going on. Everything sits on a 192.168.10.0/24 network.

Since I’ve installed the S1500 I’ve noticed some issues:

  1. My Unraid server stopped getting a static DHCP address from pfSense. It was getting a 169.254.X.X address. I did not see any entries in the pfSense DHCP log.
  2. The NFS shares to Unraid stopped mounting on my Plex server. I traced this back to NetworkManager-wait-online.service failing during boot and the server tried to mount the shares before the network was ready. I saw the DHCP request in the pfSense DHCP log.

While troubleshooting I did the following:

  1. Set a static IP address directly on my Unraid server. Worked perfectly.
  2. Manually mounted the NFS shares on my Plex server after boot. Worked perfectly.
  3. Reconnected pfSense, Unraid and Plex servers back to a dumb switch. All network issues were resolved - e.g. Unraid got its static DHCP IP & Plex server booted with no errors; NFS shares mounted as expected.
  4. Disconnected all devices from the S1500 except pfSense and Unraid. Unraid did not get its static DHCP IP address. e.g. it got a 169.254.X.X address.

It seems like the S1500 is causing some sort of issue with the DHCP server running on pfSense. I am wondering what I need to try next?

Well I answered my own question tonight after some additional digging both online and in my brain. It has been over 10 years since I’ve worked in a network role and most of that knowledge was purged years ago.

Spanning Tree was causing issues. I had to set the portfast command on the ports where my Unraid and Plex servers are connected. Otherwise the 30 second block was affecting the initial network connection.

The commands were a bit more complicated in ArubaOS that Cisco but not difficult. Create a new MSTP (multiple spanning tree profile) with portfast enabled and then assign that profile to the specific switch ports where it is needed.

#interface-profile mstp-profile servers_portfast
#portfast

#interface gigabitethernet 0/0/2
#mstp-profile servers_portfast

Rebooted both servers and they came up with DHCP static IP addresses. And they came up quicker because they were not waiting for the STP block to clear!

I’ll leave this up for reference.

1 Like