LearnDaily Operations
A Five-Minute SQL Server Morning Health Check
1 min readUpdated 6 Nov 2026
You do not need an hour to know whether the environment is fine. These checks answer the only question that matters before ten o'clock: did anything break overnight?
The five-minute order
- 1.Services and instances: is everything running, and did anything restart?
- 2.Agent jobs: which failed, and which ran far longer than usual?
- 3.Backups: is every database covered inside the target window?
- 4.Space: are any drives or log files close to full?
- 5.Errors: what is new in the error log since yesterday?
Then look for what changed
Most incidents start with a change, not a failure. Permissions, configuration values, job schedules, deployments and index maintenance all count. If something is slow today and nothing changed, you are looking for hardware or data growth. If something changed, you are looking for the change.
- Logins, role membership and permission grants.
- Configuration values and instance restarts.
- Job schedule edits and deployment activity.
Make the check repeatable
A checklist you run once is a guess. A checklist that runs every morning and shows the same numbers in the same place lets you spot the difference, which is the whole point.
The check should end with a short list of things needing attention, not a pile of graphs.
What to remember
- Services, jobs, backups, space, error log — in that order.
- Ask what changed before asking what is broken.
- Run the same check the same way every morning.
- Finish with a list of actions, not charts.
