(RAID controller) Unified way to get alarms

Hello,

I've noticed each PCI Raid card (for example ICP-Vortex ,LSILogic or ServerRAID) with their own driver uses a specific way to get alarms.

We could have several types of alarm or warning, depending on the event (degraded RAID, bad sectors detected, temperature sensor, etc...) :

- audible alarm ==> the more classical signal
- messages in kernel logs ==> not all drivers support this and each one send different messages
- messages in console ==> idem
- informations in /proc/scsi/... (for 2.4 kernel) ==> for example in /proc/scsi/gdth/0 with ICP controllers

- others ?

Unlike proprietary OS with proprietary (manufacter) driver, we have the ability under linux and with opensource drivers to use an unified way to get warning and alarms from a RAID controller.

Why don't use it ?

It will permit a single tool to manage these events without writing one per driver...

Fabien SALVI