Notifications
Who OrkStage notifies about deployments, incidents, maintenance and domains, through which channels, and how each person chooses what reaches them.
Channels
| Channel | How |
|---|---|
| In the console (the bell in the header) | Always available |
| SMTP or Amazon SES, through an outbox with retries and a per person rate limit | |
| Webhooks | HTTP POST of each event as JSON, signed with HMAC |
| Slack | A message to a Slack incoming webhook |
| Microsoft Teams | A message to a Teams incoming webhook or workflow |
Slack and Teams are webhook endpoints with another format. Tenant owners and admins add them under
Integrations, choose Send as Slack message or Microsoft Teams message, paste the incoming
webhook URL and pick the events. The message carries the title, a short text and a link to the
console. The URL must be one of the service's own hosts (hooks.slack.com for Slack;
webhook.office.com, logic.azure.com or powerplatform.com for Teams). These messages are not
signed, because Slack and Teams do not check signatures; the retries, the test button and the
automatic disabling after repeated failures work as for other webhooks.
Who receives what
Notifications go to the tenant's owners and admins and to platform owners and admins. Examples:
- a deployment started, succeeded or failed, or a production deployment waits for approval;
- an instance became degraded, a critical service turned unhealthy, or drift was found;
- an incident was opened, updated or resolved;
- maintenance was scheduled, started or completed;
- a custom domain was verified or could not be verified, its certificate was issued, or it expires in 30, 14, 7 or 1 days;
- someone tried to create an instance beyond the tenant's instance limit.
Each notification in the console links to its deployment, instance, incident or page. Mark it as
read from the bell. The matching API routes are GET /notifications and
POST /notifications/:id/read.
People who follow a status page get their own emails about incidents and maintenance, see status pages.
Notification settings
Notification settings in the console menu lists every event with two switches: Console
and Email. Everything is on until you turn it off, and your choices apply to you only. The API
is GET and PUT /me/notification-preferences:
{ "preferences": [{ "event": "deployment.started", "channel": "email", "enabled": false }] }Slack and Teams messages go to channels of the whole tenant, so they are set per endpoint under Integrations, not per person.
The worker queues emails in an outbox and retries a failed one later, up to 5 attempts in total.
| Variable | Meaning |
|---|---|
EMAIL_TRANSPORT | smtp (default) or ses |
SMTP_HOST, SMTP_PORT, SMTP_SECURE, SMTP_USER, SMTP_PASSWORD | The SMTP server, when the transport is smtp |
EMAIL_FROM | The sender address; with SES it must be a verified identity |
EMAIL_RATE_LIMIT_PER_HOUR | Emails one person receives per hour at most (default 30); the rest stay in the console |
With ses, the worker sends through Amazon SES v2 with its own AWS credentials and AWS_REGION,
never with the role of a cloud target. In the local stack, emails land in Mailpit at
http://localhost:8025.
Status page and incidents
How OrkStage derives instance health, shows it on a status page per tenant, opens incidents from health checks and metrics, and posts incident updates.
Templates and the visual builder
Create, clone, validate and publish OrkStage templates in the console, in the YAML editor or the drag and drop builder that edits the same document.