A legacy server failure at a UK ingot manufacturer turned into a two day rebuild and around £140,000 in lost production, after their only backup turned out to be cloud-only. It is one of the clearest examples we have seen of why backup and disaster recovery are not the same thing, and why the difference matters most on the day you need it.
Why cloud-only backup is not the same as disaster recovery
A cloud backup protects a copy of your data, but on its own it does not give you somewhere to run that data from. If the server holding it fails, the backup has to be pulled back down and installed onto new or spare hardware before anyone can use it again. For a business running production off that server, that gap between "the backup exists" and "the system is working again" is where the real cost sits.
What actually happened
The manufacturer was running a legacy server, in service since 2012, backing up directly to the cloud. When the server failed, there was no local hardware ready to take over. EPX IT had to retrieve the cloud backup and rebuild it onto a spare server, then reinstall and reconfigure everything until the system was working properly again. That process took two days from failure to full recovery.
The true cost
Two days without a working server meant two days of disrupted production, on a business that, like most manufacturers, cannot simply pause and pick up later. Once lost output, missed delivery slots and the recovery effort itself were accounted for, the total cost of those two days came to around £140,000. That is what one real failure, on one real server, actually cost.
What changed afterwards
Following this, EPX IT rolled out backup and disaster recovery (BCDR) appliances across the firms on its manufacturing IT support agreements: an on-site micro-server that holds a local backup, alongside a cloud copy that protects against losing the building itself. The combination means a single server failure can be recovered from on-site, in hours rather than days, while the cloud copy remains the safeguard against fire, flood or theft.
What this means for your factory
If your backup strategy is cloud-only, ask how long it would actually take to get a failed server working again, not just how quickly the data itself could be recovered. For most manufacturers running production, design or warehouse systems off an ageing server, that is the number that matters, and it is usually much larger than expected.
Ageing hardware is usually the thing nobody has been asked to look at, so it is worth having someone map what is actually running and where the single points of failure sit. That work sits inside IT infrastructure and network management, and where a server does need replacing it goes into the IT strategy and consultancy plan instead of arriving as an emergency.
Continuous monitoring under managed IT support then picks up the warning signs, so a failing disk raises a ticket before it stops a line. If you are not sure how your current provider would answer the two day question, you can grade your current IT provider against what good looks like.
