[OpenIndiana-discuss] Replacing an unavailable hd...

[John Doe]

I went onsite to replace the disk.
 zpool offline VOLUME c5t2d0 + unplug old, plug new
And nothing happens...
There was this in messages:
  Jun 14 02:30:12 data-4 ahci: [ID 517647 kern.warning] WARNING: ahci0: 
    watchdog port 2 satapkt 0xffffff06cd8cbc68 timed out
  Jun 14 02:30:23 data-4 ahci: [ID 860969 kern.warning] WARNING: ahci0:
    ahci_port_reset port 2 the device hardware has been initialized and the
    power-up diagnostics failed
  Jun 14 02:30:24 data-4 sata: [ID 801845 kern.info] /pci at 0,0/pci15d9,62f at 1f,2:
  Jun 14 02:30:24 data-4  SATA port 2 error
  Jun 14 02:30:24 data-4 scsi: [ID 107833 kern.warning] WARNING: /pci at 0,0/pci15d9,62f at 1f,2/disk at 2,0 (sd3):
  Jun 14 02:30:24 data-4  SYNCHRONIZE CACHE command failed (5)
  Jun 14 02:30:25 data-4 sata: [ID 801845 kern.info] /pci at 0,0/pci15d9,62f at 1f,2:
  Jun 14 02:30:25 data-4  SATA port 2 error
  ...
I tried:
  cfgadm -cconnect xxx and get:
  cfgadm: Insufficient condition
Nothing new in /var/adm/messages.
cfgadm still says
  sata0/2  sata-port  disconnected unconfigured failed
I guess "sata-port"
So I guess this sata port is bad (or just disabled?)...
And of course, that's one of these servers that must stays on 24h/24...
Is there a way to make the kernel try to reinitialize this sata port appart from a reboot?