Hey Community 👋
I have been thinking about how Enhance rolls it updates out for a while now and wanted to get some other views or opinions about what could go wrong and how it could be avoided or at least planned for. Don't get me wrong, I love how the updates are all centralised but I also like to be prepared for worst case scenarios.
We have come from a cPanel background where the server updates, minor and major, would not all happen at the same time - they would be phased over sometimes days or weeks and any problems picked up could normally be mitigated before they affected the next server.
With Enhance though, we hit a button and it can potentially update every single server in the Cluster and I'll be honest it aways makes me nervous hitting that button.
Yes, we have backups and we even have backups of those backups but it's more the potential disruption and client kick back that would happen if an update broke all of the web servers or database systems.
I know Enhance themselves test their updates and they advise that people also perform their own tests before updating but in reality we tend to watch the forum posts to see if everyone is updating ok before we hit our button.
If an update did somehow break the websites, due to a container issue or any 1 thing that could go wrong, is there a way to roll back the containers to it's previous version.
Is it best practice to have a number of Enhance Clusters instead of 1 big one with every server in it?
How do you currently mitigate against a failed Cluster update?
And thoughts and suggestions are welcome 🤔
Mark Donne
