Sundar,
Take a look at this site for explanations (thank google!):
Iostat -E, whats that all about? - The dot in ... --- ...
Basically, whenever I see >10 hard errors on any disk in my environment, I check with our hardware support vendor, and usually have the disks replaced. If not right-away, it is a matter of time before these disks die on you.
Soft errors = recoverable errors, related to SCSI protocol I think
Hard errors = non-recoverable errors
Transport errors = bad cabling or controller on disk
HTH.