Quick question about my upgrade process before I attempt it.
So, I am planning during the week to:
Then on the weekend when I am ready to cut:
Is there anything that I will lose on the subscribers by doing this?
I recall on 6.x I tried to achieve the same thing, however unless you have the Existing Licenses it would NOT allow you to delete the subscribers from the Clone Pub.
Not sure on the 8.x if it would allow or not, so just watch out for that gotcha.
Secondly, make sure all the phones have the latest / upgraded firmware in place on the existing cluster, so that during the cutover, all the phones come back and register normally, if they went for firmware upgrade and if you encounter an issue, it will be time consuming to know if it is a firmware upgrade issue or it is the new cluster upgrade issue.
i did upgrade and migrate from 8.6 on MCS to 10.5.2 on UCS server using PCD by changing the nodes IP address.
no need for isolated environment and phones were registered to new cluster of you change the option 150 in DHCP.
since you are changing IP address you need to perform the bulk certificate management on both cluster.
This doesn't work for me since I need to old IPs back. CIPC depends on the IPs being the same. Also, I don't use Prime.
How do I perform the bulk certificate management? Is it an installable thing or do I need TAC for that? Can it be done preventively so I won't have a problem? I did run into this on my last upgrade and needed TAC, but would like to avoid that again if I can.
Below is the procedure for the Bulk certificate management.
Refer the page # 113 and 114.
This is required if you performed migration using PCD with ip address changed on the destination node.
PCD is a good option
However you could go for second option which I have tried and works well.
Take a DRS Backup of your current CUCM which is on MCS
Fresh install same 8.x on UCS
Upgrade to 10.x
Note: Downtime is required
When underlying OS needs to be changed, cop file is needed first to install to upgrade OS before proceeding with normal upgrade
Also you could refer this documents.
Thare are at least three relevent documents:
Also, since 10.0 and beyond is only supported in a virtualized deployment, ensure that you understand all of the requirements on that front:
I'm not sure if I am reading this correctly, but are you saying that I can just do an in place upgrade by following this procedure?