Technical Maintenance Plan

System / service: [Name of the system this plan covers] Owner / team: [Accountable team or individual] Version: [v1.0] Last reviewed: [Date] Next review: [Date + 90 days]

---

1. Scope and systems

[List every component this plan covers and, just as importantly, what it does not.]

In scope

• [Application / service 1]
• [Database / data store]
• [Background jobs / queues]
• [Infrastructure: hosting, network, certificates]

Out of scope

• [Anything explicitly handled elsewhere, e.g. third-party SaaS managed by the vendor]

2. Routine task schedule

[The recurring work. Every task has a frequency and a single named owner.]

Task	Frequency	Owner
[Review monitoring dashboards and open alerts]	[Daily]	[Name / role]
[Apply routine OS and runtime updates]	[Monthly]	[Name / role]
[Verify a backup by restoring it]	[Monthly]	[Name / role]
[Update third-party dependencies]	[Monthly]	[Name / role]
[Rotate credentials and API keys]	[Quarterly]	[Name / role]
[Review TLS certificate expiry]	[Quarterly]	[Name / role]
[Capacity and cost review]	[Quarterly]	[Name / role]
[Review and update this plan]	[Quarterly]	[Name / role]

3. Monitoring and alerting

[What is watched, the threshold that fires an alert, and where the alert is delivered.]

Signal	Threshold	Alert channel
[Service uptime / health check]	[Two consecutive failed checks]	[On-call pager]
[Error rate]	[Above 2 percent of requests over 5 minutes]	[On-call pager]
[Response latency]	[p95 above 800 ms for 10 minutes]	[Team chat]
[Disk usage]	[Above 80 percent]	[Team chat]
[Certificate expiry]	[Fewer than 21 days remaining]	[Email + ticket]

4. Patching and update cadence

• Routine patches — [Operating system and runtime updates batched into the monthly maintenance window.]
• Dependency updates — [Reviewed monthly; minor and patch versions applied, majors planned separately.]
• Critical security patches — [Applied out of band within (e.g.) 48 hours of disclosure, bypassing the normal window. Define the trigger and the approver.]
• End-of-life tracking — [List runtimes/platforms with their support end dates and the planned upgrade.]

5. Backup verification

• What is backed up: [Databases, file storage, configuration, secrets store.]
• Frequency: [e.g. continuous or daily snapshots, retained for N days.]
• Restore test: [How often a backup is actually restored to a scratch environment to prove it works, and who signs off on the result.]
• Recovery targets: [Reference the disaster recovery plan for RPO / RTO commitments.]

6. Maintenance windows and communication

• Standard window: [e.g. Sundays 02:00–04:00 in the service's primary timezone, low-traffic period.]
• Notice given: [How far ahead users and stakeholders are told, and through which channel.]
• Status updates: [Where status is posted during and after the window — status page, chat, email.]
• Emergency changes: [How urgent work outside the window is approved and communicated.]

7. Service levels and response targets

Severity	Definition	Response target	Resolution / update target
[SEV-1]	[Full outage or data loss]	[15 minutes]	[Hourly updates until resolved]
[SEV-2]	[Major feature degraded]	[1 hour]	[Same business day]
[SEV-3]	[Minor / cosmetic issue]	[1 business day]	[Next maintenance window]

Availability commitment: [e.g. 99.9 percent monthly uptime for the core service.]

---

Optional sections

• On-call rotation: [Who carries the pager and how escalation works after hours.]
• Runbook index: [Links to step-by-step procedures for the most common operational tasks.]
• Change log: [A short history of major maintenance changes and plan revisions.]