03-30-2026 10:10 AM - edited 03-30-2026 10:21 AM
Hello,
I have tested both EXOS and VOSS with ZTP+ in XIQ Site Engine. Both use DHCP to learn the domain and get an IP. They both use DNS to resolve extremcontrol.local.com and they appear in the discovered window in SE.
After they are automatically moved to the site devices location following ZTP+ pending configuration changed I noticed a difference in terms of management IPs.
With EXOS the DHCP IP is converted to a statically configured IP for the Default VLAN as shown by show config command.
With VOSS the DHCP IP is kept and used on the VOSS switch. The VOSS switch is using 8.6 and I'm usig the OOB management interface. This is also withing GNS3 and it appears to prefer the OOB over the VLAN 4048 and I cannot stop this behaviour in the VM.
My question is this, is it normal to convert the DHCP IP to a static IP for EXOS or should VOSS convert the DHCP IP to a static address?
I'm also thinking we would not run into issues with other switches using DHCP and getting the same IP as another switch?
Thanks,
Rob
Solved! Go to Solution.
3 weeks ago
Yes, what you're seeing is normal. Once ZTP+ is done, EXOS usually turns the DHCP-assigned address into a static IP on the default VLAN. This makes sure that the device always has the same management address. On the other hand, VOSS keeps using the DHCP-assigned IP, especially when the OOB management interface is on. That's why you see that preference in your GNS3 setup. This difference is intentional, and while DHCP reuse could cause problems in theory, in practice site engines and DHCP scopes are set up to avoid giving out the same IP address twice.
3 weeks ago
Yes, what you're seeing is normal. Once ZTP+ is done, EXOS usually turns the DHCP-assigned address into a static IP on the default VLAN. This makes sure that the device always has the same management address. On the other hand, VOSS keeps using the DHCP-assigned IP, especially when the OOB management interface is on. That's why you see that preference in your GNS3 setup. This difference is intentional, and while DHCP reuse could cause problems in theory, in practice site engines and DHCP scopes are set up to avoid giving out the same IP address twice.
4 weeks ago
Hi Robert, we'd really need to do a packet capture with you to see where the DHCP is going when the devices are on site so we can confirm your old settings aren't getting overwritten somewhere further upstream. If you haven't already, please open a case with our GTAC team so we can walk you through running a packet capture and then interpret it with you.
4 weeks ago
Hi Sam,
I have since discovered that XIQ-SE can change the management address when onboarding under ZTP Device Defaults for the site which looks like it can allocate IPs to new switches and avoid clashes as it is aware of what other devices have been assigned IPs in the site. Will test it out.
I believe I can use onboarding VLAN 4048 and I-SID with DHCP to assign IP initially and then change to one in a Switch-Mgmt VLAN which is more static.
Thanks,
Rob