isochroma

Member
  • Content Count

    4
  • Joined

  • Last visited

Community Reputation

0 Neutral

About isochroma

  • Rank
    Member
  1. isochroma

    Internal additional Sata spin down

    Spindown support is controller-dependant, at least the type that is activated by Windows / BIOS. Most SATA controllers do NOT support spindown commands. Please see this thread for much more info: Controllers: Spindown Support
  2. Indeed it is, at least so far that I've experienced. I've had a 320GB drive running for about 5 months now under medium/heavy 24/7 load. It is very cool running (20-21 Centigrade) due to being run in a ultracooled basement server. 3 bad sectors showed up in the event viewer (eventvwr.msc, check it once a week) about 4 weeks ago. This is perfectly normal for any drive. I really like the fact that this drive isn't hiding any bad sectors, but I am a bit surprised that its sparing table is already full. The filesystem will easily handle any more bad sectors. Certainly I'm excited to find out how fast corruption really happens and since the drive can't lie (or hang) after the sparing table is full like others do, I will be informed of these events.
  3. Modern file systems are more than up to the task of managing bad sectors. The MFT/FAT can hold an arbitrary number of bad sector remaps, unlike the very limited physical remapping present in the drive hardware/platter design. Also, using a sparing table in hardware means the drive head is always seeking to the beginning (or end) of the platter surface whenever it hits a bad sector, in order to read/write a sector in the 'spare' section, interrupting smooth transfers and wearing down the actuator mechanism. If the bad sectors are dealt with in software, there is no excessive seeking or interrupted transfer because that sector is simply never used.
  4. Why would you want to disable the feature? If after 7 seconds the drive can't read/write the sector, it should report the bad sector to the controller, which will report it to your OS. In turn, your OS will mark the sector bad in the MFT/FAT, then it won't be used anymore. This is how things should work on ALL drives. But instead, most drives try to hide their bad sectors by 'relocating' the reference to an available 'sparing sector'. Once the small part of the drive reserved for 'spare' sectors is full, then if it's a WD non-TLER drive it will hang the machine because it cannot relocate and will not report. If it's a non-WD drive it will probably report. Key point: if drives reported all bad sectors from the beginning, you would have an idea how quickly corruption is occuring. Your software would be able to avoid sectors near bad ones, etc. By the time the defect table is full, your drive is in serious trouble. Only so many skeletons can be hidden in the closet. The ugly truth is this: modern drives develop bad sectors EXTREMELY FAST, due to their high density. To cover this up, manufacturers have uniformly implemented 'defect management', aka. lying to the host controller about spreading bad sectors.