In a tax office’s file server, four hard disks work as an array so that the failure of a single one does no harm. One fails in spring, and the warning goes to a mailbox nobody reads any more. In autumn a second disk fails, of all times while the array is rebuilding. The data is lost. The backup job had also been reporting errors for weeks, to the same address. Hardware will fail, that much is certain. The only open questions are when, and whether you are prepared.
Malfunctioning hardware is a technical risk source. In its definition, ISO/IEC 27005 lists technical sources as a type of their own alongside human and environmental ones. There is no intent, and yet the consequences of a failure can weigh just as heavily as those of an attack.
What is behind it?
This covers all physical components: servers, storage, network devices, workstations, printers, uninterruptible power supplies, air conditioning units, controllers and connected devices. Each of these components has a limited lifespan, and many depend on one another. If the air conditioning fails, the servers follow; if a core switch fails, the whole network is affected.
How failures arise
- Early failures. New devices with manufacturing defects often fail within the first weeks.
- Wear. Fans, hard disks, batteries and capacitors age, and the failure rate rises over the years.
- Creeping faults. Defective memory or a faulty controller corrupts data without any system crashing.
- Firmware bugs. The software inside the devices can itself contain bugs that only appear under certain conditions.
- Overload and under-sizing. Full storage, power supplies that are too small or weak cooling push devices to their limits.
- End of support. For old devices there are no more spare parts and no more security updates.
Impact
Availability comes first: systems stand still until a replacement arrives. With storage faults, integrity suffers, and without a working backup data loss looms. Defective batteries and power supplies can even start fires. Failures of components meant as a safeguard are particularly treacherous, such as the battery of the uninterruptible power supply that only gives out during a power cut.
Practical examples
Aged batteries. The batteries of an uninterruptible power supply are seven years old. During a power cut of a few seconds, the unit fails to switch over to battery and shuts down itself. All servers go down hard, and afterwards a database only starts with damaged tables.
One air conditioner, no reserve. At the weekend, the only air conditioning unit in a laboratory’s server room fails. There is no temperature sensor with alerting. On Monday two servers and a storage system have been damaged by overheating.
The old storage on the network. A network storage device for which the vendor has not shipped updates for years is still running in the office of a trade business. Attackers take it over through a known gap and use it for attacks on others. The organization only finds out through a complaint from its internet provider.
Typical threats
These BSI elementary threats stem from this risk source:
- G 0.25 Failure of Devices or Systems and G 0.26 Malfunction of Devices or Systems: the core of this risk source.
- G 0.45 Data Loss: defective storage media and controllers.
- G 0.8 Failure or Disruption of the Power Supply, G 0.9 Failure or Disruption of Communication Networks and G 0.10 Failure or Disruption of Supply Networks: when power supply, network equipment or cooling are themselves faulty.
- G 0.27 Lack of Resources: full or undersized components.
- G 0.12 Electromagnetic Interference: faulty devices disturbing others.
- G 0.1 Fire: overheated batteries, power supplies and cables.
- G 0.47 Harmful Side Effects of IT-Supported Attacks: outdated devices that attackers turn against third parties.
Effective controls
Prevention:
- A.7.13 Equipment Maintenance: maintenance according to manufacturer instructions, scheduled replacement of batteries and fans.
- A.7.11 Supporting Utilities: power supply and cooling with reserve capacity and regular tests.
- A.8.6 Capacity Management: watching utilization and expanding in good time.
- A.8.14 Redundancy of Information Processing Facilities: duplicate design where the tolerable downtime calls for it.
Detection:
- A.8.16 Monitoring Activities: monitoring the condition of storage media, temperature, batteries and network devices and routing warnings to a responsible person.
Response:
- A.8.13 Information Backup: tested backups that can be restored onto replacement hardware.
- A.5.30 ICT Readiness for Business Continuity: spare devices, recovery plans and practised procedures.
Risk sources in ISO 27005
The definition in ISO/IEC 27005 recognizes three types of risk sources, and the technical one is among them. In the list of typical threats in Table A.10, technical failures form a group of their own, including the failure of a device or system, which the standard classifies there as accidental. Among the aspects for likelihood criteria, it explicitly lists technology failure, and when assessing accidental sources it names influences on equipment malfunction. In Annex A.2.2 the standard recommends recording dependencies between assets, since a fault in a supporting asset such as a server affects the business assets that rely on it. Finally, the treatment options include removing the risk source, for example by replacing a device before the end of its service life (→ ISO/IEC 27005 Information Security Risk Management).
Sources
- ISO/IEC 27005:2022, definitions 3.1.6 and 3.2.7, clauses 6.4.3.3 and 7.3.3, Annex A.2.2 and A.2.5.1
- BSI IT-Grundschutz Compendium: elementary threats G 0.1 to G 0.47
- ISO/IEC 27002:2022, clauses 5.30, 7.11, 7.13, 8.6, 8.13, 8.14 and 8.16