What logic did you use to determine which disk to pull? Your append seems to indicate they were selected at random (try one, then another - that's not good as you mess up the last access times and further degrade the array...)

To really know which disk is which, just before powering down, simplest is to run "smartctl -a /dev/sdx | grep Serial" - where sdx is the disk in question. then check the disks for that serial number... alternatively "hdparm -I /dev/sdx | grep 'Serial\ Number' " will also provide the same information.

Note that swapping disks may change the way linux enumerates the disk logical numbers, and understand that the disk number and "RaidDevice" numbers in the output from mdadm bear no direct relationship with physical slots or bays.

Do we assume from your error that the system is no longer bootable? If so, you will need to boot a live linux system on CD, DVD, SD card or USB drive to find out the state of your raid array - impossible to know from the information rovided whether it can be salvaged...

Raid has saved data here several times - but you need to understand totally what you are doing, not panic, and ask questions first before touching anything if not 100% sure...