Work

Ops

Nine servers watched from one phone, with the routine repairs done without me.

Internal tool

Problem

Nine servers, and the honest question is not whether they are up but whether the thing that watches them is still alive.

Solution

One view over nine servers with 36 watchdogs and a dead-man switch that reports when the monitor itself dies. A native Android client delivers the alerts as push.

How it is used

  1. 01Open the phone and see all nine at once.
  2. 02An alert arrives as a push, not as an email nobody opens.
  3. 03If the monitor goes quiet, the dead-man switch says so.

Results

  • 9 servers under 36 watchdogs.
  • Native Android client with push alerts.
Access
Internal tool
Kind
Infrastructure

Stack

  • Kotlin/Compose
  • Node
  • Docker

Want to talk about it?