This is a cool article I found that studied whether there was any correlation between hard drive temps and failure rates. The conclusion? There is none. Which I personally found surprising.
Their testing methods seem severely flawed. They aren't taking into consideration the absolute extreme for HDD temperature. My honest opinion? 30-34*C isn't hot. At least not hot enough to do damage to the micro controller on the HDD controller PCB. Shit, most HDDs operate at that temperature normally. If they removed their cooling fans from their data center or turned the AC off, you'd see a different story. I can, without a shadow of a doubt, assure you that you would see drives reaching in the 60*C+ range. This might pose a potential risk to drive failure on the micro controller attached to the controller PCB.