On 14/09/16 19:16, Daniel Sanabria wrote: > Other than replacing the green drives with something more > suitable (any suggestions are welcome) WD Reds or Seagate NAS. I don't think they make them any more, but Seagate Constellations are fine too. My Toshiba 2TB 2.5" laptop drive would be fine. The tl;dr version of the problem with Greens (and any other desktop drive for that matter), if you haven't read it up yet, is that when the kernel requests a read from a dodgy drive, it just sits there, *unresponsive*, until the read succeeds or the drive times out. And the drive will time out in its own good time. If the kernel times out *before* the drive, and by default the kernel does so after 7 secs, while the drive can take two minutes or more, then the kernel will recreate the missing block and try to write it. The drive is unresponsive, the write times out, and the kernel assumes the drive is dead and kicks it from the array. That's why you need to increase the kernel timeout, because you can't reduce the drive timeout, and which is why a flaky hard drive will cause system response to fall off horrendously. Cheers, Wol -- To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
