Scheduled maintenance on your status page
Deploying database migrations or swapping load balancers should not look like a mysterious outage. Scheduled maintenance on your status page sets expectations: we know, we planned it, here is the window.
StillOnline lets you publish incidents (including maintenance) on the project timeline. External HTTP checks may still run — plan messaging so customers are not surprised when the page turns degraded or down.
Quick answer
Post scheduled maintenance on your StillOnline status page before deploys so customers see a planned window — not a mystery outage. Probes may still flip yellow or red; explain that in the incident body and keep owner Telegram alerts on for surprises. Mark Resolved when checks are green and note what changed.
Before maintenance starts
Announcing early means deploy night is about communication you control, not surprise tickets.
- Announce early — create an incident post: title “Scheduled maintenance”, status Scheduled or In progress when you begin.
- Use the same public URL you share with customers — public status page guide.
- Notify owners — your team still wants Telegram or email when checks fail unexpectedly. Connect the StillOnline bot in settings; Telegram alerts. Distinguish planned downtime (your post) from accidental down (probe alert).
During the window
The status indicator color and your customer-facing words can disagree — align them on purpose.
| Approach | When |
|---|---|
| Leave checks running | Short deploy; expect brief degraded/down — explain in the incident body |
| Point check to maintenance page | Returns 200 with “under maintenance” if you control the route |
| Brief real downtime | Honest red on page matches reality — incident text is critical |
Template copy for the incident body: incident post template.
After maintenance
Close the incident so history shows you finished on time.
- Mark incident Resolved with end time.
- Confirm checks green on dashboard and public page.
- Optional: note what changed (“API v2 live”) — builds trust.
Subscribers who used Subscribe with Google on the public page get email on status changes tied to incidents.
Cadence for indie teams
Match ritual to how often you ship.
- Weekly deploys: one-line maintenance post becomes habit.
- Rare big migrations: schedule 24h ahead in status plus email stakeholders the link.
- No page yet: five-minute path → project → first maintenance post same day.
Related guides
FAQ
Will StillOnline Telegram alerts fire during planned maintenance?
If your health URL returns non-200 during deploy, StillOnline probes may alert your owner channel — brief alerts are normal. Explain planned work in the maintenance incident; do not monitor staging on the same check if noise hurts. Telegram owner alerts.
Can StillOnline maintenance posts be automated from CI?
On Pro, use the incident API to open and close maintenance from pipelines — REST service uptime. You still own the customer-facing wording in each update.
Are StillOnline incidents available on the Free plan?
Publishing incidents on the status timeline is supported; verify current limits on Pricing. A maintenance post costs less than a support avalanche during a deploy.
How is StillOnline maintenance different from a degraded probe?
Maintenance is communication on the timeline; probe color is measurement from HTTP. Align both when you expect real downtime — incident post template.
Should I post StillOnline maintenance before every weekly deploy?
For B2B customers, a one-line scheduled post builds predictability. For silent deploys with zero user impact, a short maintenance post is still cheaper than surprise tickets — public status page guide.
Do StillOnline subscribers get email when I publish maintenance?
On a public page, Subscribe with Google subscribers receive email when incident status changes — separate from owner alerts in settings.