I have a situation, Where I have to perform software/ Hardware Upgrade of Cisco 6500 series switch.
The reason for Upgrade:- Because existing software/Hardware is already EoS/EoL.
slot#1 - WS-X6724-SFP 24 Port
slot#2 - WS-X6708-10GE 10gig 8 Ports
slot#3 - WS-X6708-10GE 10gig 8 Ports
slot#4 - Empty slot#5 - WS Sup 720-3B
slot#6 - Empty
slot#7 - WS-SVC-FWM-1
slot#1 - C6800-32P 10G 48Ports
slot#2 - Empty
slot#3 - Empty
slot#4 - Empty slot#5 - VS-SUP2T-10G
slot#6 - Empty
slot#7 - Empty
Current IOS: 12.1
Targeted IOS for new Sup2T 15.2.1-Sy7(MD)
Since such migrations need a complete switch shutdown followed by removing the old hardware and install the new hardware(OK understood cabling migration is already in place).
Now my aim is to achieve this task with minimum downtime/cutover/services impact time in place. So here is my thought to reduce the downtime.
Considering that I have only one physical chassis, and based on above-provided details under Current Hardware I have slot#6 free in hand. So can I install the new Sup2t card on slot#6 and let it load and come online. Then I'll get the console access of that Sup card and by using an additional Flash card, I'll install and upgrade the software version of Sup as per targeted 15.2.1-Sy7(MD) version.
Once the software is loaded, I'll remove the Sup from slot#6 as this step was done to avoid software upgrade on doom day(migration day).
Now my understanding is, I can achieve the above software upgrade task without interrupting the already running the current process on old sup and modules.
So on migration day I'll simply shut down the whole chassis, remove the old modules and sup, install the new module and sup card, power-on the chassis, upload the new configuration and BINGOOOO I don't need to upgrade the IOS version because that was done as part of pre-activity homework.
personally this bit risky and you may have longer down time something go wrong playing with Live chasis insert supervisors.
I had same issue, the whole chasis went offline while i was inserting supervisor, cisco TAC mentioned after long investigation, it one time incident. ( crash logs not helped much).
i had serious downtime for an hour or so.
i suggest to go with new chasis parallel to the setup and test and migrate one one by one link, so your setup side by side old and new together. so you have minimal down time, and you have option to reloback if something not working on new chasis.
Do you have QoS in place ? make it change to new QoS format MQC and MLS