Load balancing by default in Enhance?

Subhan

I came across a technique online for distributing traffic to a WordPress website across multiple servers, using load balancing.

This method sends incoming traffic to a group of back-end servers, and if one server goes down, traffic is automatically redirected to the remaining servers.

Does Enhance have a similar feature built-in, or do we need to set up load balancing manually to achieve this?

In other words, can Enhance automatically distribute traffic across multiple servers, or do we need to configure it separately?

Thanks

grradmin

No, Enhance does not offer this feature and don't think it will for a very long time if at all.

enhost

Load balancing is an awesome tool when it comes to improving reliability, I would say its not really the target feature set for a hosting panel as sites which would benefit from it are probably hosted on more specialist panels.

That being said there is nothing stopping you running a load balancer infront of two enhance servers, you would need to keep the servers up to date, ensure they are using the same database, and object storage ideally. Disable SSL on the enhance servers and let the load balancer handle it.

You could use traefik, or caddy as a load balancer

wav3front

I believe Enhance should at least support automatic failover.

XN-Matt

wav3front This would at least require some sort of live storage replication to a designated second host (which you'll know will be up at the time the first is down) or shared storage....

twest

You can get automatic recovery using AWS EC2 instances:
https://aws.amazon.com/about-aws/whats-new/2022/03/amazon-ec2-default-automatic-recovery/

Downside is it's very expensive while performance is pretty low.

I can't imagine Enhance gets such capabilities in the near future, maybe something in 5-10+ year range. I like the way it is now with decommissioning and recommissioning servers, simplifies the recovery process by letting you roll out downed sites onto a new server with just a few button clicks. Being a manual process of initiating the recovery, it gives a human administrator the final word to go ahead with it or not. Besides false positives, one may prefer not recovering if their datacenter techs are able to quickly diagnose the issue and replace the hardware or fix the network issue - whatever it may be, fixing the server offers some benefits over going for a recovery.

We keep a spare server running idle just for this situation. It's fully setup and configured/optimized, ready to load sites in event of catastrophe. Having this "standby server" sit idle should cut down recovery times at least 15-30 minutes because the server is already provisioned/running. Of course, if a server dies we'll go into DR mode, first assessing the situation, troubleshooting what's wrong, then making the judgement call whether to let DC techs work on the server or go for recovery using Enhance's recovery option. We run Enhance backups every 4 hours to make sure backups are very fresh, minimal data loss in a disaster.