IDCF Cloud ransomware: in the cloud is not backed up
Ransomware encrypted servers for 495 customers of a Japanese cloud, and backup data could be retrieved for 206. Why one copy belongs somewhere else.
Published · 3 min read

At about 3:40 a.m. on 7 October, ransomware stopped and encrypted virtual servers on IDCF Cloud, the infrastructure service run by Japan's IDC Frontier, affecting 495 companies and local governments. The attacker came in through a flaw in open-source software in the login system of the portal customers use to run their servers, then used the management console to take administrator rights over the virtualisation platform itself. By IDC Frontier's own account, restoring those customers' systems where they stood looked difficult, and by 20 October it could retrieve backup data for 206 of the 495. Being in the cloud, it turned out, was not the same as being backed up.
The company says it had no contact with the attacker and paid no ransom, and its report of 27 October found that no data on customers' servers had been leaked. A message left on customers' consoles made claims of its own, BleepingComputer reported: a breach in seven minutes, 239 hypervisors reached, 554,153 snapshots wiped. Those are the attacker's numbers, and the company's report does not repeat them. The report does show where recovery came from. The first 194 customers whose backup data it could retrieve, confirmed on 14 October, had used the platform's own snapshot feature, which saves the state of a server's storage at a point in time.
A snapshot lives where the server lives
A snapshot is a good defence against the everyday failures: a bad update, a deleted folder, a change that has to be undone by lunch. But it sits on the same platform, under the same administrator rights, as the server it protects. When those rights are what the attacker took, whether the snapshot survived is something the provider has to find out, and IDC Frontier's customers waited a week for the first answer and nearly two for the second. The company's president wrote that affected customers still faced "the need to rebuild their systems".
For an office here, moving to the cloud takes away the server in the back room, and much of the worry with it. It does not move the question of where the other copy is. The provider protects its platform; whether your records survive the platform's worst day depends on a copy kept beyond it.
What it means for the systems I build
The systems I build run in the client's own accounts, with the client's hosting, code and data, so the backups belong to the client too. That makes the backup question one the client should have answered at handover: not whether the data is in the cloud, but where the second copy lives, which credentials can delete it, and when someone last restored from it. The records that matter most are the ones nobody could type back in: open orders, customer histories, a year of invoices. A backup also has to be guarded as well as kept, which is the other half of the DMW-linked case.
Five questions for your own cloud
- Where is the copy that does not depend on your provider? Keep at least one backup with another provider or in a separate account, under credentials your everyday administrators cannot use to delete it.
- Does the plan rely on snapshots alone? They are quick to restore from, and they sit inside the platform they protect. Treat them as the first copy, not the last.
- When did you last restore? Restore into a fresh environment, not over the old one, and time it. Part of IDC Frontier's recovery help was moving customers to alternative environments.
- How is the provider's console protected? IDC Frontier's was entered through its customer portal's login system, and before reopening it the company is limiting access to customers' IP addresses and making multi-factor authentication mandatory. Ask whether yours does, and switch on what you can.
- Could you rebuild from your own code? If the code and the steps to set up a server are in your own repository, a lost server becomes a rebuild rather than a reconstruction.
If nobody in your office can answer the first of those questions, the fit check is a good place to start.
Sources
- IDC Frontier, IDC Frontier Reports Investigation Findings and Next Steps Regarding Unauthorized Access, 27 October 2026
- BleepingComputer, Ransomware attack disrupts Japan's IDCF Cloud used by govt clients, 8 October 2026
Enrique Lacambra III builds internal business systems from Cainta, Rizal.
Tell me what is broken and I will tell you if it is worth building. A 20-minute call, no charge, and I will say so if you do not need me.
