[CentOS] Software RAID on Centos 4 - new issue

OK, I have managed to get software RAID 1 running on my wonderful Proliant
server but now have two challenges:

1) Fresh after an OS install and reboot, the RAID array starts to do its
stuff but straight away (as part of the boot messages) I am told that the
second disk is 'not ready for command' and the system hangs as soon as
resyncing starts. I think this is fixed by adding an hdparm line to turn on
DMA (I was also losing ticks), but cannot check this because...

2) As part of the testing, I rebooted the system with the second drive
pulled out, the whole lot then boots and I can use mdadm -a to add the
second drive to the mirrored pair. The drives resync at a decent speed and
the pair works fine for the duration, but when I reboot the second drive is
not part of the mirrored pair unless I manually add it again.

Looking at the md superblock for the second disk, I am wondering if all the
attempts at setting up the RAID array have confused it as this is what I get
- the last section looks wrong (??) 

 mdadm -E /dev/hde
/dev/hde:
          Magic : a92b4efc
        Version : 00.90.00
           UUID : b0dcd24f:7044414a:1f0ba458:adabce9e
  Creation Time : Thu Sep 16 06:21:59 2004
     Raid Level : raid1
    Device Size : 193157440 (184.21 GiB 197.79 GB)
   Raid Devices : 2
  Total Devices : 2
Preferred Minor : 0

    Update Time : Wed May 11 09:13:50 2005
          State : clean
 Active Devices : 1
Working Devices : 2
 Failed Devices : 1
  Spare Devices : 1
       Checksum : 2652883e - correct
         Events : 0.15217


      Number   Major   Minor   RaidDevice State
this     2      33        0        2      spare   /dev/hde
   0     0      22        3        0      active sync   /dev/hdc3
   1     1       0        0        1      faulty removed
   2     2      33        0        2      spare   /dev/hde


Any suggestions on getting this pair talking properly?

Thanks

Nigel