Upgrade procedure overview
Follow the upgrade procedure for your situation within a maintenance window because network availability cannot be guaranteed. Plan to have at least a two-hour maintenance window in case problems arise. Some of the preparation can be done before the upgrade to minimize the downtime.
Rolling upgrade for switches
To minimize the amount of traffic down on a network, the rolling upgrade procedure involves upgrading each of the switches serially. This procedure takes longer overall but it ensures the least amount of downtime for the traffic on the network. The rolling upgrade procedure is useful for network environments that require live traffic running during their maintenance window.
The process first involves the upgrade of one switch. Second, a subsequent transition of the traffic is moved over to that upgraded switch. Third, the remaining switch is upgraded. While the first switch is upgraded to this release, the other switch running version 10.00 still uses MCLAG until the VSX configurations are made to the first switch. Once the VSX configuration is applied to the first switch, the downstream MCLAG interfaces become active approximately 10 minutes after the configuration is applied. The remaining device can also now be upgraded to this release.
Upgrading both switches at the same time
The procedure for upgrading both switches at the same time is useful when you want the shortest amount of overall time to upgrade. This procedure is also useful when you can afford to have both switches offline during their maintenance window. This procedure uploads the software on both switches and boots them both into the new software at the same time. This procedure lets both switches boot in parallel, which saves time.
Network traffic will NOT be forwarded across the switches until VSX is configured on the upgraded switches.