on 01-29-2016 09:32 AM
Here's a scenario:
1) Mesh link quits working (building to building)..
2) APs and antennas mounted on masts on building roofs. I can physically get to the mesh point's console port.
3) Mesh point has lost its configuration and just needs to be reprovisioned without removing it from the mount to move it to another building (to be able to get a wired ethernet connection to the controller).
1) Is it possible to put enough configuration on the mesh point via the console port environment variables to get it to re-establish the mesh link with the portal? I have the mesh profile PSK from the controller.
2) If so, what are the minimum environment variables needed to be applied to make this happen?
3) If not, is there a method (cellular hotspot & RAP/IAP, etc) to get the mesh point a virtual connection to the controller, so it can be reprovisioned without unmounting it?
on 01-29-2016 03:10 PM
1) Not really no. There are some settings written on to the flash of the AP that have to be done by the controller and are not able to be input via the console.
3) what I recommend, and what I usually do, is take another AP configured as a mesh point with the wired port enabled to allow a wired device to connect to the ethernet port of the temporary mesh point, and stand that device up as a point to the portal (put the temp point in the same group, etc as the one you're replacing). Wire the down/wiped point to the temp point, purge it and let it come up as it would if it were a CAP. Log in to the controller (via the wired connection or via wireless from the temp point), provision the down point, reload it, and disconnect it from the temp point and power down the temp point. You should be able to watch the previously down point come up.
Your other option, if you have a spare controller, is to make the spare controller a 'Local Controller' to your master so that it gets the exact config, mesh recovery profile, etc. Then take that local controller out to the down AP, enable a DHCP server on the local, let te down mesh point come up on the local controller as a CAP, provision it in to the same group and give it the same name. Reload, disconenct from the temp controller, and watch the previously down point come up. We've used this for larger scale issues before (usually when the installer unknowingly mounts them all before getting them configured back in the day).
Both methods work pretty well and are pretty easy. It sounds complicated, but it's really not. You can email me at jerrod.howard -@- hpe.com and I can likely send you one of the cheat sheets we made.
Sr. Techical Marketing Engineer