Wireless Access

I have 1 3000 series controller on 6.4.2. I have 4 new 7205 controllers now. I need to migrate the 3000 series to a 7205, then add redundancy. The 2 other 7205 will be redundant masters. What order do you recomemnd the migration take place? 

 

also, need to keep controllers on 6.4.2 to accomidate legacy WAPs. 

For starters the 7205 controller minimum ArubaOS is 6.4.3.0.

Why do you have 4 controllers? What design have you used to determine that you need 4?

2 controllers to replace the local 3000 controller and 2 master controllers in the data center to support them and others. We have ap105 the reason for the legacy OS.



Sent from my Verizon, Samsung Galaxy smartphone

Also can we create a master cluster across datacenters over a L3 link? 

Master redundancy is L2/VRRP based only. 

 

Coming soon with 8.x and Mobility Master will be the capability to have redundant L3 Mobility Masters. 

Ok, going back to the original questions. What order would you plan this controller deployment? example, deploy the masters first, confgure the 3000 to use the masters, then migrate the local controllers or? 

 

Also will 6.4.3.0. suppot WAP 105s? I take it I need to upgrade the 3000 to 6.4.3.0. and demote the 7205 to 6.4.3.0. then backup and migrate configs? 

1. Back up everything, move backups to laptop or anything 

2. Upgrade your current deployment to 6.4.3, validate everything is working

3. Bring up your 7205, at least one, and downgrade to 6.4.3

4. Make the 7205 a backup master to your 3000 controller (this will replicate the config). Once everything is synced, take the 7205 off and bring it up on a table/bench

5. Change the 7205 VRRP priority to be the value you want for the master and change the IP address to be the master IP address

6. Bring up your other two 7205s as locals and make the third your master backup (or whatever you're doing with them).

7. Change the config to point the LMS IP to the new 7205 locals (VRRP IP usually)

8. Testing everything on the bench first, bring up an AP on the 7205s to make sure everythign works.

9. Take your outage, power down your 3000 (but leave in place in case you need to roll back), bring up your 7205s on the production network. APs should come back up ( as your master is the same IP address)

10. Assuming everything was done correctly, all yoru APs should come back up on your new controllers, pointed to the new local VRRP IP, and be working without issue

 

This assumes that you do the part of getting the network cofig (vlans, port assignments, etc) all done correctly on the initial setup on your bench/lab table. 

This is great. How much down time should I anticipate/estimate for this project start to finish? 

One last question here. I will setup the new local 7205 controllers to run 6.4.3 to accomidate the legacy WAPs. Can I run the Master 7205 cluster on a higher version of AOS or do they need to be downgraded to 6.4.3 also? 

 

 

Jerrod, also, on step 4, setting up the 7205 to be backup master to the 3000, is there any gotchas related to the different port numbering on the 7205 that needs to be addressed? 

??