Sometimes the Solaris Solstice Disk Suite (SDS) can do really strange things. Below you see a mirror where both submirrors are in state Needs maintenance.

d0: Mirror
Submirror 0: d10
State: Needs maintenance
Submirror 1: d20
State: Needs maintenance
Pass: 1
Read option: roundrobin (default)
Write option: parallel (default)
Size: 1027776 blocks

d10: Submirror of d0
State: Needs maintenance
Invoke: after replacing "Maintenance" components:
metareplace d0 c1t0d0s0 
Size: 1027776 blocks
Stripe 0:
Device Start Block Dbase State Hot Spare
c1t0d0s0 0 No Last Erred

d20: Submirror of d0
State: Needs maintenance
Invoke: metareplace d0 c1t1d0s0 
Size: 1027776 blocks
Stripe 0:
Device Start Block Dbase State Hot Spare
c1t1d0s0 0 No Maintenance

In this example submirror d20 failed and all I/O were stopped to this device. Then an error occured on submirror d10 causing this device going to Last Erred state but it is still working. To get out of this situation you have to do an metareplace -e $SUBMIRROR $DEVICE on the submirror in state Maintenance. Applied to the example above you should do an

metareplace -e d0 c1t1d0s0

After the syncronisation is finished you have to do the same with the second submirror:

metareplace -e d0 c1t0d0s0

Done and don’t forget to validate your data.