[PC-BSD Testing] Mount tray detects SCSI disks as USB devices
ken at pcbsd.org
Wed Aug 28 05:53:53 PDT 2013
On 08/27/2013 14:13, Ken Moore wrote:
> On 08/27/2013 10:39, Ken Moore wrote:
>> On 08/27/2013 07:32, Claudio L. wrote:
>>> One correction:
>>> On 08/25/2013 12:46, Claudio L. wrote:
>>>>> da1 is listed as a "USB Device", and then it fails to mount (tries
>>>>> to mount with mount_msdosfs). This is particularly weird, since
>>>>> da0 is not shown, only its slices, but da1 shows as a separate disk.
>>>>> da1s1 is also listed as "USB Device" (even though it's my root ZFS
>>>>> disk already mounted).
>>>>> da2 is correctly listed under its Windows label
>>> I recently swapped the cables and got confused when writing the
>>> post, so my root ZFS drive is actually da2, not da1 (sorry, my
>>> mistake), and da2 is not shown, so the paragraph above should've said:
>>> da1 is listed as USB Device (even though it shouldn't be listed).
>>> da1s1 is correctly listed under its Windows label.
>>> da2 is not shown at all (which is correct).
>> That is good to hear, at least the current OS pruning is working
>> properly.. :-)
>> If I recall correctly, if a parent device (da1) has children (da1s1)
>> the algorithm will only list the parent of it satisfies a stricter
>> set of criteria (seeing if it has additional available space/
>> filesystems). I will check on that though and make sure. I will
>> probably just need to tighten up the restrictions on parent devices
>> even further.
> I just fixed the issue with parent/child devices, it looks like there
> was a typo (or a later change to one of the subfunctions) that was
> only looking for 2+ children to ignore the parent instead of any
After some addition back-and-forth between me and claudio off-list, the
USB/SCSI distinction should now be fixed in the mount tray source. This
has also resulted in a better default naming convention for USB/SCSI
devices (based on the hardware identification/vendor) than the previous
~~ Ken Moore ~~
More information about the Testing