Server Maintenance Checklist
A monthly server maintenance checklist for Linux and Windows: disk, logs, backups, certificates, accounts and monitoring. Schedule it and keep a record.
- Typical schedule monthly
- The checklist 5 sections, 20 items
- Variables
{{month}}
Use this checklist for the routine monthly look at your servers: the small checks that prevent the classic avoidable outages, such as a full disk, an expired certificate or a backup job that quietly stopped weeks ago. It suits small teams running a handful of Linux or Windows servers, virtual machines or cloud instances without a dedicated operations team. Patching has its own checklist. This one covers everything around it.
Who owns it
One named person owns the monthly run, usually whoever is responsible for infrastructure that month. They do not have to log in to every server by hand. Dashboards and reports are fine, as long as someone actually looks at them and acts on what they find.
Common pitfalls
- Trusting alerts that never fire. An alert that has been silent for a year may be broken. Check that monitoring still covers every server and that test alerts arrive.
- "The backup job is green." A job can succeed while backing up the wrong path or an empty database. Look at backup sizes and do a small restore now and then.
- Ignoring slow trends. Disk usage that grows a little each month is easy to miss in a daily view. Compare with last month.
- Accounts nobody owns. Old SSH keys and local admin accounts from former staff or contractors pile up quietly.
Schedule the run in Runbook for the first working day of each month. The history then shows each completed month, which helps with audits and handovers.
The checklist: Monthly server maintenance {{month}}
Health and capacity
-
Disk usage checked on every server, and anything above your threshold cleaned up or ticketed
Compare with last month to spot steady growth before it becomes an outage.
-
Memory, CPU and load trends reviewed for the past month
-
Inode usage checked on Linux servers with many small files optional
-
Failed services and units checked
Linux: systemctl --failed. Windows: services set to automatic that are not running.
Logs and monitoring
-
System and application logs scanned for new recurring errors
-
Log rotation and retention working, and old logs not filling disks
-
Monitoring covers every server in the inventory, with no hosts missing
-
Test alert sent and received by the on-call person
Backups
-
All backup jobs succeeded in the past month
-
Backup sizes look plausible compared with the source data
-
One file or database restored as a spot check optional
Security and access
-
Pending OS and package updates reviewed against the patching plan
-
User accounts and SSH keys reviewed, and leavers removed
-
Sudo or local administrator membership checked
-
TLS certificates expiring within the next 30 days identified and renewal scheduled
-
Firewall rules reviewed for temporary openings that were never closed optional
Housekeeping
-
Clock synchronisation (NTP) working on all servers
-
Old snapshots, temporary files and unused images cleaned up
-
Server inventory and documentation updated with any changes
-
Findings for
{{month}}logged as tickets with owners
Checklist as Markdown
Frequently asked questions
What does server maintenance include?
Routine server maintenance covers capacity checks, log and monitoring review, backup verification, security hygiene such as accounts and certificates, and housekeeping like time sync and cleanup. Patching is part of maintenance too, but many teams run it as a separate cycle because it needs maintenance windows.
How often should server maintenance be done?
A monthly cycle works for most small teams, with automated monitoring covering the time in between. Some checks, like backup job status, are worth looking at daily or through alerts. Others, like a full access review, can be quarterly.
Can I use this checklist for Windows servers?
Yes. The checks are the same in principle; only the tools differ. Use Event Viewer instead of journald, Task Scheduler instead of cron, and local groups instead of sudoers. Add notes to the items with the commands your team uses.
Updated