Following on from today’s multiple outages at FidoNet I thought it might be worth while detailing a little on where we host our servers and why.
As many who read this will know, I founded FidoNet back in 2001 after spending years at both Demon Internet (1992-1999) and Redbus Interhouse PLC (1999-2001). I then sold FidoNet and tried to retire back in 2017.
After failing at retirement and discovering just how bad most of the competition was, I set up FXRM to focus on quality hosting. I needed to host a few virtual machines initially for my own projects and FidoNet were kind enough to offer me a deal, in exchange for some continued technical support on my part. In addition to FidoNet’s hosting, FXRM also utilises Droplets with Digital Ocean, Virtual Servers with Hostinger, and more. The beauty of no longer owning the infrastructure is we can focus on operating the services. Virtual Servers are pretty much a commodity these days, whether it’s Azure, AWS, GCP, Hostinger, Digital Ocean, OVH, etc .. we can spin up containers and add them to our distributed cluster based on where it makes sense to host machines, and without the headache of maintaining multi-million-pound networks. The downside, as we discovered this morning, is that when one of their engineers does something (or fails to do something), it can have a knock-on effect.
In realistic terms, FidoNet has been absolutely amazing and incredibly robust. They have continued to invest in hardware and infrastructure well beyond what I could have done at the time, and they have focused on building out the backbone.
Yes, it means I feel hamstrung when an update goes awry and I ask “why on earth were you even contemplating that at noon on a work day” but at the same point, the changes being made should not have had an impact, and it is quite possible that the server would have failed and rebooted with or without them being logged into it at the time.
The beauty of where we are currently is that we are able to host on multiple platforms simultaneously and not have to invest in the manpower to maintain these backends, nor do we have to service the finance bills or worry about scaling operations to pay the bills. Instead, we’re focusing on delivering quality support, simple-to-use control panel interfaces and providing an above-par service level.
At the same time, we’re still working with FidoNet to ensure that they have the benefit of our 30+ years of experience with infrastructure and systems, and catching problems before they arise (most of the time) ensuring the minimum of disruption to service.
As FXRM grows, we may choose to implement our own infrastructure again in certain places, or we may continue to decide to leverage the best-in-class HyperVisor technologies (or whatever comes along next) with the knowledge that we’re spreading our risk across multiple providers to ensure the minimum of disruption and the lowest latency we can to our customer base. We make the technical decisions so you don’t have to.
Should you have any questions or want to discuss this in more detail (or implement a similar solution for your business needs) I remain available to answer questions through the support system.
Kind Regards
Jon Morby
Managing Director
FXRM Services Ltd
