[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: how to test disk for bad sector



Thank Marco!
i have installed debian to other disk of same PC, no problem.
anyway i try 2nd command you suggest, see below for result

[liveuser@localhost-live ~]$ sudo dd if=/dev/zero of=/dev/sda1 bs=8M status=progress

We trust you have received the usual lecture from the local System
Administrator. It usually boils down to these three things:

    #1) Respect the privacy of others.
    #2) Think before you type.
    #3) With great power comes great responsibility.

14386462720 bytes (14 GB, 13 GiB) copied, 151.001 s, 95.3 MB/s
dd: error writing '/dev/sda1': No space left on device
1719+0 records in
1718+0 records out
14418915840 bytes (14 GB, 13 GiB) copied, 153.319 s, 94.0 MB/s




On Saturday, August 29, 2020, 12:41:55 PM EDT, Marco Möller <talby@debianlists.mobilxpress.net> wrote:


On 29.08.20 14:20, Long Wind wrote:

> it pass write test!  i really don't know where installation go wrong.


After the disk itself seems to not cause the problem, at least not
because of bad blocks, confirm that the data highways on your
motherboard and any involved data caches are all still operating fine.

(1) Read at highest possible speed from the device from which during
installation is read from by directing the read data to the device
"/dev/zero" ("/dev/zero" is a device which is not doing anything with
received input, all received data is immediately after reception simply
ignored = forgotten = thrown away):
  sudo dd if=/dev/DeviceReadingFrom of=/dev/zero bs=8M status=progress

(2) Write at highest possible speed a huge amount of data to the disk in
order to check if the path to the disk also works free of errors under
high load, and likewise also check that the cache inside the disk still
performs free of error, by writing data received from the device
"/dev/zero" (you will destroy all data on the disk, this begins at the
beginning of the disk and will thus first erase a present partition
table; the disk will after this test have to be partitioned and
filesystems have to be newly created; no old data can be expected to be
recoverable from the disk) (in this case, from the device "/dev/zero"
there will be received values "0" as long as requested and thus the
disks becomes overwritten with zeros):
  sudo dd if=/dev/zero of=/dev/WritingToThisDisk bs=8M status=progress
If you do not cancel this process by STRG+C before, then it will end
with an error message when the disk is full and you can see from the
last reported status how much data has been written to the disk.

(3) Do the same as above but involving the full data highway from your
device you are reading from to the device you are writing to:
  sudo dd if=/dev/DeviceReadingFrom of=/dev/WritingToThisDisk bs=8M
status=progress

If all these tests run through without errors (with test (2) running
until the disk is full), then I would say that it is safe to assume that
it is not a hardware problem of disks, disk controllers or motherboard.
My only left assumptions would then be, that maybe in the BIOS some
security feature is activated which interferes with the installation
process, or some other peace of hardware lets the system crash when
becoming addressed during the installation process, like some
incompatible graphics board or some other incompatible peripheral
component.

Good Luck!
Marco



Reply to: