This issue has severe gameplay consequences and it keeps popping up for months. Every once in a while, especially in a raid fight, FPS may be frozen for 4 to 10 seconds and then it goes back to normal (it may occur once in 10-20 minutes or so). There is disk activity at the time this happens (or at least the disk led is on). I had thought it's a sleeping function of the hard disk and attempting to remove all such functionality did not help (I did not discount any hard-coded sleeping functionality on the device though). For a while, I believed it's overheating since reducing the frame rate of the game manually, which reduces CPU-use, seemed to help but later I noticed the issue popping up again as normal (vents are cleaned with pressed air). There is a high probability, without discounting anything, that GPU temperature is not the issue since it appears modern enough to be comfortable on 40FPS I was forcing it to operate. The main beef I have with it is that with all my experience in computing, I can't exactly pin point an FPS freeze (not drop) in relation to disk activity. Asking at sites such as SuperUser did not appear to produce any targeted answers either. I was hopping a WoW player may have more specific information. Lately I attempted to disable Recount since I heard it is heavy on resources but while I believed it did work at first, the issue popped up again.
Addons running are autolagtolerance, autolog, bagnon, chronobars, combustionhelper, cooldowns, dbm, dominos, icehud, magemanabar, magenuggets, omen, pawn, powerauras, prat, raidachiv, raidbufstatus, rangedisplay, rawr, tinydps, webeburnin, xperl.
PC is a
HP dv7 Pavillion
Lately I suspected its the saving of a large (combat) log file, but deleting it it had the issue occurring normally with only 40mb length of log.
While I was investigating disk device issues I had suspected some "Reallocated Sectors Count" on S.M.A.R.T. I have not concluded on that but the number is stable for months, and most probably since the purchase.
The disk included in that particular PC had similar issues on Macs it had appeared. However the Firmware those used to fix the issue appears to be older than the one included on the one at hand (so there is high probability without discounting anything that it is not related).