This message is from the T13 list server.
Great suggestion with respects to the HDD defective sectors! Possibly the HDD could publish the number of sectors reserved for reassignment and then the number of sectors that have been reassigned. Percentages can be derived from those 2 pieces of information. Regards, Nate -----Original Message----- From: sraposo [mailto:[EMAIL PROTECTED]] Sent: Friday, March 22, 2002 2:35 AM To: [EMAIL PROTECTED] Subject: [t13] Hdd end of life This message is from the T13 list server. My purpose is to post two suggestions that, if a T13 member considers them such an interesting ideas that encourages him to formally propose them, could make part of a future AT/A version. It's based on HDD problems that occur everyday. It's a point of view of someone who deals with real life HDD issues. Indeed they are features or an extention of SMART feature. HDD has a pool of spare sectors to replace those defective ones that pops during HDD life. The amount of spare sectors is vendor-specific and I guess that it is not T13 business how many sectors a HDD maker reserves to perform as spare. Since this pool is limited, there will be a moment that this spare sector will run out and HDD will become, who knows?, nearer to its end. Anyway, user has no means to know if a HDD has run out of spare sectors or if its sectors pool is full or empty. So, a flag to be added to the word 93/94 or some other word of Identify Device command result (maybe SMART result?) would indicate that this pool is empty or not. Maybe other flags could be added to give a more detailed status of this pool like a 100% full flag, 50% full flag, 25% full flag, 10% full flag etc. There is another very commom trouble that perhaps is related to the problem above, but causes a very, very, very disastrous effect: the HDD startup process, that critical period of time that determines the HDD sentence: life or death. My suggestion is to reserve a bit at word 93/94 or some other to indicate that there were problems or some error (problably a correctible one) during the read of a firmware table data or code at this startup process. Maybe a specific bit could be set when the last firmware table/piece of code backup is put on service. I hope that the utility of both suggestions is clear. As some people on this forum clearly demonstrate, HDD are made to attend people's need, and people need safety, pople's data need safety. I apologize for this long e-mail and I hope I could give some contribution. Regards, Sergio Raposo - Brazil
