Yesterday, I received a phone call.
It was from a client I worked with years ago, but I had lost contact with them in early 2020. Back then, I was told about an ongoing migration “to the cloud.” I gave my opinions (they have poor connectivity and a high need for reliability without significant investment, since they are a healthcare facility, and dozens of people, besides patients, would be stuck if services became unreachable). In 2015, I had set up twin local servers for them, syncing the VMs every 5 minutes to be ready to switch without the full burden of true HA. The setup was based on Proxmox, with daily backups to another location.
Yesterday morning, the internal person responsible for this called me in a panic. In one phone call, I discovered that:
- They are still using that server - without any updates since 2019 (Proxmox 5.x).
- They removed the twin server because, according to this person, it was “useless.”
- They are using a Windows 2008r2 VPS in production as a main server and a VPS with Windows XP (!!!) to manage some external operations. This VPS “exploded” due to an error by one of their staff (!!!), and they don’t know if there are any backups or how to restore them.
None of these machines are accessible externally, but all are on a LAN with hundreds of devices. All on the same LAN.
I (politely) exploded, having been convinced for years that this server had been decommissioned. I had already stressed in 2020 the need to update the underlying infrastructure, explaining that the physical server dates back to 2015, that you can’t run such old software on Windows 2008r2 and Windows XP, and that Proxmox 5 has been EOL for ages. But they responded, “we are allocating funds for the new cloud setup,” even though their connectivity is unstable and the costs would be absurd.
I decided to help them anyway because I was thinking about the dozens of sick people stuck and waiting for services to be restored. I connected via VPN they provided, and luckily, I discovered that the backups were still happening daily (just as I had set them up years ago and last checked in 2019) and that they were intact. I restored the (small) Windows XP VM from the backup of the previous day, and they resumed working.
After resolving the emergency, I called back and explained that such a situation is inconceivable. They replied that they are still “planning the migration to the new cloud system” but need to adjust connectivity, configurations, software, VPN, etc. After 5 years. At that point, I lost my patience and, courteously but firmly, emphasized that while they are “dreaming” of a “cloud” they will never achieve (and have already paid tens of thousands for “design”), their entire operation runs on a stable setup that just needs to be updated. They have 8 locations, and now I discover they’re all like this. At any moment, an unsolvable problem could arise. But no use: all their IT funds are diverted to the “study and realization of the new cloud solution” - which, by their own admission, won’t be ready anytime soon. Maybe in a couple of years, maybe. I tried to alert the owners as well, but they insisted, “the project has started, a lot has been spent, and it cannot be stopped. And there are no funds for any further spending on the current infrastructure, not even 500 euros.”
I explained that, given their internal and external situation (non-excellent and unstable connectivity, without any prospect of improvement, especially for 2 sites), they need a local copy of the data, processed locally, and then, for security, always copied externally. But to no avail. Their “cloud consultant” has been “working hard” for years to migrate everything, but there are “many issues,” and it will take “time and money.” And unfortunately, they are following that path.
I wonder when we will realize that blindly following trends and hype can be (besides unnecessarily expensive) truly, truly dangerous.