Back to the day job today (yes I know it’s Saturday – but some things can only be done on a weekend!), so no study update. However, here I am upgrading CUCM (major version upgrade), and I am working with three remote locations on high latency links to a centralised deployment. To cater for this our phone firmware is loaded on IOS gateways in the locations. This is accessed thanks to the load server parameter on the phone device configuration. This serves two purposes: 1. It causes the phones to load firmware locally and; 2. It prevents the phone from attempting to upgrade over the WAN if the firmware is not on the load server (handy if you want to add a newer firmware than provided by the upgrade – which you need to do after the upgrade).
The problem with this setup is, what happens when all the phones reset due to the switchover of CUCMs after the switch version ? Well it seems that about 10 phones upgrade from the gateway and the rest re-register. I’m not sure if this is a hard limit within IOS, but I have consistently seen that through this upgrade – so here I am resetting the device pools to ensure that all the devices are eventually upgraded (and there’s another 10 phones done!).
Load servers are still a great tool in a small deployment – but maybe I should look more closely at peer-firmware upgrading (which seems like magic). Oh yes – and it takes about 7 minutes to upgrade each of the 10 phone batches!
No comments:
Post a Comment