On Sat, 25 Jun 2016, Mayuresh wrote: > I tried setting the apm value to 255 (and also experimented with 254, > 253). The Load_Cycle_Count rise rate came slightly down with this, though > it still kept increasing at least by a couple of hundred daily.
A hundred head unloads/day is probably ok and within the life expectancy of most proper laptop HDDs (which should be rated for 300k head unloads on the typical laptop HDD, and 600k head unloads on the very good ones (e.g. Hitachi/HGST)). It is like writing ~1Gbyte/day on a good SSD: stop worring about it and enjoy the device, it will eventually break down because of some other reason anyway. > while [ true ]; do date; smartctl -a /dev/sda | grep Load_Cycle_Count; > sleep 30; done > /tmp/hdtick > > sleep 30 was arrived at with trial error (sleep 60) was found to be > insufficient. Seems right. > It stopped increasing Load_Cycle_Count, though I am curious whether this > is safe to do. Hope I haven't found just another way of reducing the life > of the drive! It is unsafe in that a mechanical shock will cause head damage far more easily, as they will never be safely in the unload area. Also, you are increasing "head flying hours" in place of unloading them, so you are trading one kind of mechanical stress for another. Also, ensure you let that disk spin down at least once every 48H. If it is not unloading heads, it is not spinning down either. > Also, curious why apm setting did not solve the problem in the first > place. Because laptop HDDs are not supposed to leave heir heads flying when idle for long in the first place (or, for that matter, spinning), as impacts and sudden acceleration are a far greater danger on the typical use of a laptop. Some HDDs are far better adjusted / engineered than others, though. -- "One disk to rule them all, One disk to find them. One disk to bring them all and in the darkness grind them. In the Land of Redmond where the shadows lie." -- The Silicon Valley Tarot Henrique Holschuh
