Angry work venting [1/?]
All of the websites I'm responsible for are down right now and there's nothing I can do about it.
Our load balancer started dropping connections. We put in a ticket; no response for 2 hours. Now the LB is failing to load our most popular ecom site, so I call in and try to get someone to do something about the problem.
Shortly after that, all of our sites went down, hard.
It's been 6 hours since we reported this and the problem is now WORSE THAN WHEN WE REPORTED IT.
Angry work venting [2/?]
The only ticket update I've gotten that wasn't "I've escalated this" or "I will get back to you soon" was the following, verbatim:
"We still investigating. We have determinate that the Loadbalancer is working correctly. We need to check if there is an issue with the VMs."
We tested that. Extensively. I swiched hosts to point directly at the webservers and they all work great! Point at the LB, they all time out. I pointed this out earlier in the ticket, too.
Angry work venting [4/4]
I already contacted our infrastructure team and gotten them to repoint DNS to the webservers directly, bypassing the hosting provider's shitty LB. A single server might not handle the traffic by itself, but at least it will handle _some_ of the traffic while our Hosting provider figures out how to get their thumbs out of their assholes.
I can't even begin to describe how angry I am with our current provider. This isn't even the worst thing they've done to us.
Angry work venting [4/4]
@literorrery For the time being we've got enough horsepower on each of these webservers to handle 2x average peak traffic. I'm pretty sure we'll be alright, but round-robin is a solution we might look into if that changes. Thanks for the advice. ♥️
Angry work venting [4/4]
@mawr Round-robin DNS will solve that in the short-term. It's not as efficient as a LB, but it will get you around having to have one and distribute load around all your webservers instead of putting it all on one that might go down.