Short answer: upgrade when the supported path, application checks, recovery access, and maintenance window are ready—not simply because a new LTS exists. Keep urgent security updates moving while you decide.
What should you check first?
Confirm the current release, provider image support, third-party repositories, backup recoverability, and a console or out-of-band recovery path. A VPS upgrade without recovery access turns an ordinary compatibility problem into a lockout.
When is upgrading reasonable?
Upgrade sooner when the host is lightly customized, the provider documents the path, services have a staging check, and you can schedule a maintenance window. Delay when the machine runs unsupported packages, has no tested restore, or is carrying an irreplaceable workload.
A small-VPS decision matrix
- Low customization: standardize, snapshot, validate, then schedule.
- Many external repositories: inventory and replace or disable them before upgrading.
- Customer-facing service: test the application and rollback plan first.
- Security emergency: apply the vendor fix; do not wait for a distribution upgrade.
Read the Ubuntu kernel update triage guide and broken SSH recovery guide before choosing a date.