Prevent mdadm soft-raid to mount degraded?

[Blindleistung]

Hi Community.

I was wondering if someone knew how to set up the linux softraid 'mdadm' in a way that it would only auto-mount arrays if all disks are present.

The default configuration is to auto-scan all partitions for raid-headers, and assemble and start any found. Even degraded.

I could not find a hint in man or forum pages...

The point is avoiding a resync if one disks just happens not to start properly. Also to make sure I don't miss a fault. So if a disks fails, I would notice it, solve it, and mount manually.

I tested the default behavior:

- disconnect on of six disks in a raid6 array.
- start machine. It automounts degraded.
- power off, reconnect disk.
- start machine. It remains degraded. Disk is not
added back into the array.
- Add manually with --manage /dev/... --add /dev/...
- array resyncs.

Any help?

[Rubberman]

Well, that is the PURPOSE of RAID 5+ arrays - to allow you to continue to operate when one drive has failed (degraded mode). This allows you to replace the failed drive while still being in operation (if in a performance-degraded manner). If you don't need that capability, and are only interested in striping to get maximum performance from your drives, then you need to use a different RAID setting. The mdadm tool should allow you to do that, but remember that if one drive fails, you lose EVERYTHING...