Topic: Than now under Windows to search/isolate for defective drives?
Somehow so it was added that I never had to isolate the defective units on HDD - the last time I did it on diskettes, and HDD to me came across either completely workers, or [almost] completely dead where there was no sense to potter with insulation. And here it was formed 320-gigabajtnyj Hitachi Z5K500 on which defective all some hundreds megabytes (files/sektory are not read and are not written), but the reserve of hardware changeover is already settled (on the fuss beginning there was: Reallocation Event Count - 70/0x347, Current Pending Sector Count - 5/0x960). I want to isolate the remained defects and to give a disk to children under storage of films for the TV set. Connected it through the USB-adapter to a working notebook with Win7 and launched normal (not fast) formatting. It crackled and clicked a disk of hour two then informed that defects it is not revealed. I strongly was surprised and launched besides native scanning from Local Disk / Tools - it too worked a pair of clocks, with similar result. Thought that in the course of formatting the disk managed to correct defects, launched hd_scan - is not present, all dips and reading errors on places. The Microsoft writes to documentation that chkdsk with a key/b searches for the defective units and eliminates them from among the free. Launched under Win7 - finished minutes through ten, informing "no problms, 0 bad blocks". Loaded in Win 10.0.16299.248, switched a disk there, launched chkdsk/b - worked minutes fifteen, with similar result. Looked in Task Manager at a state chkdsk in the course of operation - and at it any of exchange counters (read/write/other) does not grow. That is, it not only readings/records, but also even IOCTL does not fulfill. Opened disk exchanges in Resource Monitor - there too practically the silence, pity background tens kilobyte per second. Actively exchanges with a disk chkdsk only right at the beginning when scans the auxiliary areas , and in the end, after end of "scanning" of the free areas. It turns out, what standard means in such cases are absolutely useless not only for an exception of defects, but also even for their detection? But on flash cards they somehow find defects? In what a difference? Downloaded with HGST their utility Windows Drive Fitness Test (WinDFT) is any sad poverty. In archive lies unpacked EXE, its one more copy lies in the form of the installer where the empty help-file which "is shown" in a window at pushing button Help is enclosed to it. If to mark a disk in and to push Scan - a daw in it is dropped, happens nothing. At double to a clique window SMART, where all entirely green, including Current Pending Sector Count opens. Long Test in some seconds falls out with Failure - similar, its unique task is determination of presence of errors, anything more. Personally I would feel ashamed such poverty to distribute from an official site even to the small company, not to mention large, but at them, obviously, other criteria. Macrorit Disk Scanner 4.2.0 something finds out, though activity indication is made awfully - when disk operations for a long time , not clearly, whether the disk brakes, whether application . This scanner has a possibility to continue from the specified place, yes here the trouble - it anywhere does not show, on what sector stopped. It is necessary to calculate sector on percent, just in case moving back with overlapping. Before it it is necessary also to restart the scanner as repeated he considers scanning of the same disk "multi-disk scan", and it is accessible only in the Pro-version. After that Current Pending Sector Count fell to 1, absolute value increased to 0x15c0. Old HD Tune 2.55 managed to scan for two hours all disk, counting nearby 1 % of the defective sectors, but showed errors only there where already at all were not read. There, where read operation was completed though in a minute, but is successful, it shows suitable areas, though it in a root incorrectly. To sense from this scanning again any as found out it only showed on the schedule, but did not register in and did not deduce the list. HDDScan 4.0, all such stylish, in mode Verify flew by on a disk with any extreme speed (about 800 Mb per second), showing the schedule corresponding ideally serviceable HDD, and without finding out any error. The disk thus did not publish a sound. I and did not understand, in what sense of this mode if it has no relation to reading from a surface. Short Self-test for a minute reached 70 % then produced "error with read element failed" and it was completed. Extended Self-test for the same time reached 10 % and it was completed with the same message. In mode Read creaked half-days, but made the list of sectors, both "purely erratic", and excessively long read. And on those thanks - now though it is possible to look for the utility which under this list generates requests to . A report window formed about half an hour, having inflated thus from hundred megabytes to eight hundred. The schedule turned out here such: By the way, with the interface they explicitly overdid - to cling boundaries for change of the sizes it is possible only for shades which are abroad and if the window is pressed to edge of the screen or this shade is not so visible, and to hook there is nothing. After these violations Current Pending Sector Count reached to 1/0x2b58. Whether there are utilities under Win32 or though under , able competently (not only on error status code, but also on command runtime) to define defects to do it not by sectors, and more large-size blocks, and at once to isolate their means ? Or now it is accepted, what at presence even unique unrecoverable sector the drive should be thrown out? Well and a twist of fate: this disk has been bought eight years ago for disk changeover in the notebook which has fallen from desktop. The old disk clicked and showed lowering of attributes in SMART, but after in USB-boxing till now works as storage of animated cartoons, despite repeated copying of a fair part of contents.