On 2021-11-12 at 16:52, Gene Heskett wrote:
> On Friday 12 November 2021 15:12:15 Dan Ritter wrote:
>
>> Gene Heskett wrote:
>>> That will be good, but getting rid of the first raid10 I built
>>> is needing tactical nukes. Its taking almost 40 minutes a drive
>>> to zero them and start over. And this machine is acting like an
>>> 8086 machine doing it. very very slow. gkrellm is showing all 6
>>> core in bright orange. Not any great temp rises though, staying
>>> below 35C for all 6 cores. The heat sink/radiator is huge, so
>>> huge I can't put the side panel back on the tower. 5" fan is
>>> turning silently at maybe 400 revs. I think I overbuilt it ;o)
>>
>> No need to do it the hard way:
>>
>> For each disk, run
>>
>> # wipefs /dev/sdX
>>
>> which will not destroy anything ; it will list the commands needed
>> to remove the existing filesystems.
>>
>> Then run that command, generally of the form wipdefs -o 0x1000 or
>> such, which will complete in seconds.
>>
>> Repeat for the next disk.
>>
>> -dsr-
>
> root@coyote:~$ wipefs /dev/sde -o 0x1000
> wipefs: error: /dev/sde: probing initialization failed: Device or
> resource busy
> And I have to reboot as mdadm has no stop command. Thats BS.
Eh?
From the mdadm man page (at least on my machine):
-S, --stop
deactivate array, releasing all resources.
And from the EXAMPLES section:
mdadm --stop --scan
This will shut down all arrays that can be shut down (i.e. are
not cur‐
rently in use). This will typically go in a system shutdown script.
While it's probably not possible to stop the mdraid backend (since IIRC
it's part of the kernel?), it should certainly be possible to tell it to
let go of the devices it's managing.
--
The Wanderer
The reasonable man adapts himself to the world; the unreasonable one
persists in trying to adapt the world to himself. Therefore all
progress depends on the unreasonable man. -- George Bernard Shaw
Attachment:
signature.asc
Description: OpenPGP digital signature