Reservation for selenium concentrators

Question

I am launching a large selenium grid and trying to achieve high availability. Currently, I have one point of failure in the Hub - if this machine goes down, then my grid goes down. I want to achieve redundancy.

I am looking for approaches that people are taking to solve this problem. The hub is configured so that I cannot load the balance. Is the best I can achieve master / slave interception?

Are there any packaged solutions for this?

If not, any tips to mitigate the impact of such a failure (what logic triggers the transition to another resource), how to minimize the time of connecting nodes to a new hub?)

Thanks in advance!

Answer 1

Yes, you can use the Apache load balancer configured as a proxy server to switch to the backup master slave. It should work fine, although the configuration is a bit tough.