Hey All,
We're in the process of migrating our DHCP servers. We've rolled out four successful scope migrations, the next being a clinic that has Aruba APs.
Both DHCP servers have identical scopes, options etc. After changing the ip helper-address in the clinic routers SVI, all computers, IP Phones and access points successfully moved over and obtained IPs from the new scope after their old leases ran out.
Old scope and new scope have matching IP address designs, after the move the old scope is deactivated and the old helper address removed from routers configuration.
The four 105 model APs at this medical clinic now constantly are stuck in a rebootstrap cycle. They will obtain an IP, the after exactly 8 mins they will go offline.
"Dec 10 14:04:58 sapd[735]: <311004> <WARN> |AP Goodbys Creek 1@10.32.81.10 sapd| Missed 8 heartbeats; rebootstrapping"
After failing to the backup-lms (in this case the master) I see in our logs:
Unable to contact switch: HELLO-TIMEOUT. Last Ctrl msg: HELLO len=1134 dest=159.XXX.XX.XXX tries=10 seq=0
failed to get master-ip
(159.XXX.XX.XXX being our master VRRP address)
If I shut these ports down and point them to the old ip helper-address, the APs will come back online and stay 100% stable.
Has anyone experienced this issue? Our access points obtain connection to our master at boot through DNS. Both old and new DHCP servers point to the same DNS server where aruba-master.domainname is resolvable to the masters vrrp.
The IP address the access points pull are completely routable on campus and can be pinged the entire time, even while the heartbeats are failing. Its seems like the GRE tunnels are forming and then never getting any data across?
Any help would be much appreciated!
All controllers use matching 6.3.1.5 code.
Thanks.