Management access
The base principle
Management or administrative access must be limited to authorised authenticated users and utilise multi-factor authentication wherever possible.
Application Program Interface (API)
APIs are preferred over Secure Shell (SSH) connections, as by comparison they generally offer greater technical security limitations without the need for parsing commands.
Automated diagnostic data collection
It should be exceptional to directly administer a server/node when adequate diagnostic data collection sends underlying technical data to a place where it can be correlated and analysed.
Pre-defined, pre-audited
Tools such as Systems Manager and comparable techniques over preferred over manual intervention (such as human interaction over SSH) as the intervention path can be carefully designed to avoid human error and effectively instruct pre-audited actions to be taken on an administrator’s behalf.
Secure Shell (SSH)
Use of bastion or ‘jump’ boxes for access into systems is a useful technical security design that also helps ‘choke’ and control such sessions.
Through immutable infrastructure and server design, state-less cluster expansion/contraction and automated diagnostic data capture the need to SSH into a server/node should be increasingly less common.
It should be exceptional for an individual to login to a server/node via SSH and execute commands with elevated privileges (typically, root
).
Using SSH
SSH must be strictly controlled, and environments should be segregated so that no single bastion or ‘jump’ SSH server can access both production and non-production accounts.
SSH shells must be limited to users who need shell (by comparison to users who will use SSH as a port forwarding tunnel).
Joiners/Movers/Leavers processes must be strictly enforced (optimally, automated) on SSH servers as they are a critical and privileged access method.
SSH should not be password-based, and should use individually created and purposed SSH keypairs. Private keys must not be shared or re-used.
Contact and Feedback
For any further questions or advice relating to security, or for any feedback or suggestions for improvement, contact: security@justice.gov.uk.