On 2021-06-13 4:03 a.m., Patrick O'Callaghan wrote:
On Sat, 2021-06-12 at 16:06 -0700, Samuel Sieb wrote:
> On 2021-06-12 3:46 p.m., Patrick O'Callaghan wrote:
>> On Sat, 2021-06-12 at 14:43 -0700, Samuel Sieb wrote:
>>> On 2021-06-12 10:19 a.m., Tom Horsley wrote:
>>>> On Sat, 12 Jun 2021 10:44:19 -0600
>>>> Joe Zeff wrote:
>>>>
>>>>> According to
https://github.com/openzfs/zfs/issues/10891 has
>>>>> been
>>>>> depricated since last year. Disable and mask it and that
>>>>> should
>>>>> fix
>>>>> your issue.
>>>>
>>>> It is listed as a "static" service on my system. Can those be
>>>> disabled or masked?
>>>
>>> It can't be disabled, but it can be masked.
>>
>> I've masked it and rebooted. Makes no difference to boot time in my
>> case.
>>
>> Jun 13 00:13:59 Bree systemd[1]: Starting Wait for udev To Complete
>> Device Initialization...
>> Jun 13 00:13:59 Bree udevadm[413]: systemd-udev-settle.service is
>> deprecated. Please fix nm-initrd.service not to pull it in.
>> Jun 13 00:14:29 Bree systemd[1]: Finished Wait for udev To Complete
>> Device Initialization.
>> Jun 13 00:14:30 Bree systemd[1]: systemd-udev-settle.service:
>> Deactivated successfully.
>> Jun 13 00:14:30 Bree systemd[1]: Stopped Wait for udev To Complete
>> Device Initialization.
>
> That is weird. It shouldn't get run if it's masked. Is the service
> file in /etc/systemd/system by any chance?
$ ls /etc/systemd/system/systemd-udev-settle.service
/etc/systemd/system/systemd-udev-settle.service
I think if you use "ls -l", you'll find that it's a symlink to
/dev/null. In the other thread we figured out that it's actually in the
initrd as well, so masking it on the system only has a partial effect.