vCenter Server Reduced Downtime Upgrade
Reduced Downtime Upgrade feature was introduced in vSphere 7.0 U3. But what exactly is this new feature and how does it affect the different deployment types of vSphere?
From a supportability perspective, as of vSphere 7.0 Update 3, Reduced Downtime Upgrade (RDU) is only supported for VMC on AWS and it is also part of Project Arctic, which was announced at VMworld 2021. Find more information about Project Arctic on our blog and the Virtually Speaking Podcast. This feature helps tremendously during vCenter upgrades and significantly reduces that downtime of vCenter services for the aforementioned cloud services.
Current methods of updating/upgrading vCenter Server can result in extended periods of downtime as services have to be stopped, patches and binaries installed, scripts run, and services have to be restarted. vCenter Server Reduced Downtime Upgrade brings a migration-based upgrade approach differently from current migration approaches, where the result is a significant reduction of downtime of vCenter servers. Let’s take a look at how the current migration approach works.
The current migration process includes:
- Deploy new VCSA running the new version
- Stop source VCSA services (DOWNTIME)
- Export/Import data from source VCSA and make it compatible with the new format (DOWNTIME)
- Shutdown the Source VCSA (DOWNTIME)
- Run first boot scripts on new VCSA with the copied data (DOWNTIME)
Clearly, there are multiple steps where downtime is a factor as data is being copied over to the new target vCenter Server, and services are brought back online. Let’s compare this with the Reduced Downtime Upgrade process.
- Initialize Upgrade Process
- Stage target VCSA by deploying a new VM with the new version
- Configuration and database data is replicated from source VCSA to target VCSA
- Switchover (DOWNTIME) – promotion of target VCSA
- Clean leftover data from old VCSA
With Reduced Downtime Upgrade, it is evident how there is only a small downtime window while the switchover between the vCenter Servers occurs, resulting in minimal downtime of the vCenter availability. The biggest advantage is that the vCenter Server Appliance (VCSA) state, operating system and services are upgraded outside the downtime window.
Downtime is proportional to the stop/start of the VCSA services, while the current migration upgrade time is proportional to the time needed to copy the data from one VCSA to another.
If the upgrade operation is not completed within the expected time, the upgrade is rolled back to the previous state/VCSA.
As previously mentioned, RDU is part of VMC on AWS and Project Arctic offerings, where VMware Site Reliability Engineers (SRE) assist in the upgrade/update of your vCenters, allowing the customers to offload this task to VMware SREs and focus on other tasks within their environments. This is a great way to start migrating some of the infrastructure to the cloud without having to move workloads on Day 0.