Maintenance windows
Check what a maintenance banner or connectivity incident means for server status and alerts.
Status during maintenance
Alpacon occasionally schedules maintenance on the platform, or works through an unplanned connectivity incident. Either one can make your servers look offline for a while, even though nothing is wrong with them. This page explains what you’ll see and what happens to your alerts.
When maintenance is coming up or already in progress, a banner appears at the top of the app:
- Up to 24 hours before scheduled maintenance starts, a banner shows the start time.
- Once scheduled maintenance is in progress, an amber banner shows when it’s expected to end, or that no end time is set yet.
- During a connectivity incident on Alpacon’s side, a red banner appears. There’s nothing you need to do.
You can dismiss the banner; it comes back if a new window starts.
If the app loses contact with Alpacon while it’s open, it keeps showing the last banner it loaded, marked Last known schedule, until it can check again.
If the console can’t reach Alpacon
When Alpacon’s service in your workspace’s region isn’t responding, for example during maintenance, the console can’t load the workspace. Instead of an error about the URL or your access, it shows a page saying it can’t reach the workspace right now:
- View service status opens status.alpacax.com, where incidents are posted.
- Retry tries the page you were on again.
If this browser showed a maintenance banner for that workspace in the last few hours, the same banner appears on this page, marked Last known schedule. It’s the schedule as of your last visit in this browser, so a change made since then (an earlier end, or a cancellation) won’t show until the console can reach Alpacon again. In another browser or on another device, the page appears without a banner.
What happens to your alerts
While a maintenance window or incident is active, your servers may show as offline or stop reporting metrics, so the alerts that depend on hearing from a server are paused: the offline alert, and any monitoring rule alert that fires because a server has gone quiet rather than crossed a threshold. This is separate from the quiet period after maintenance you run yourself, such as a reboot or agent restart—see Server disconnection. Their absence isn’t a sign everything’s fine—it means the alert is being held until the window ends.
Monitoring rules that alert on a metric value (a threshold you’ve set) are not affected and keep working as normal.
Once the window ends, Alpacon checks in on your servers:
- Any server that’s still unreachable gets a single alert.
- Any server that reconnected is cleared with one recovery notice for the whole workspace, not one per server.
If an incident is raised without a set end time, Alpacon stops holding your alerts after 24 hours at most, even if the incident notice hasn’t arrived yet—so alerts can resume before you’re told the incident is over.
Email notices for incidents
An unplanned connectivity incident sends you an email when it starts and another when it’s resolved. Scheduled maintenance doesn’t send email; the banner, shown up to 24 hours ahead, is the only notice.
These emails belong to the System category in notification settings, listed there as “Platform-level notices”—the same place you manage email for any other category.