IT alerting by e-mail, SMS and WhatsApp
Server monitoring is only as good as the alert that reaches the right person. Rapitt sends each alert once, on the channels that person chose, about the clients they look after — and holds the rest until morning.
Book a demoThree channels, each with its own switch
Sent through your own SMTP server, so alerts come from your domain and pass your mail policies.
SMS
Short, plain messages for the on-call phone: host, client, what happened, when.
WhatsApp alerts for teams that already run operations there — common across the Gulf region.
Routing per subscriber
Every subscriber chooses three things: which alert kinds they want (host offline, disk filling, certificate expiring, tunnel down, security grade dropped and others), the lowest severity worth a message, and which clients they hear about.
An engineer on one account sees that account's events only. A client's own IT contact can subscribe to their organisation without ever receiving another client's alert. A manager can take critical alerts for everyone and nothing else.
Quiet hours and digests
Each person can set quiet hours. During them, info and warning alerts are held and sent as one digest when the window ends, so nobody wakes to forty messages about a disk at 81%.
Critical ignores quiet hours. A host offline, a tunnel down or a stopped service is sent at once, whatever the time.
One delivery per alert, per person, per channel
Alert fatigue comes from repetition. Rapitt sends an alert once, and does not send it again until the condition has cleared and returned. A flapping service produces one message and a recovery, not one every five minutes.
Every delivery is logged with its channel, recipient and result. A failed SMS or a rejected e-mail is visible in the console, so a silent phone is never mistaken for a quiet night.
Written to be read on a bridge call
Alerts name the thing — host, client, service, value and time — and end with what to do. The tone does not rise with severity; a critical alert is short and plain.
Last reading 03:09: CPU 12%, memory 61%. Tunnel up.
Check power and the hypervisor console.
Alerting questions
Can alerts go to a ticketing system?
Yes — through e-mail-to-ticket, or by polling the alarms route of the monitoring API from your service desk.
Are paused hosts alerted?
No. A paused host or client raises no alerts, and the console shows who paused it and when so the pause is never forgotten.