On 6/16/21 3:11 PM, Ed Greshko wrote:
On 17/06/2021 02:27, Patrick O'Callaghan wrote:
> On Wed, 2021-06-16 at 22:36 +0800, Ed Greshko wrote:
>> On 16/06/2021 21:46, Patrick O'Callaghan wrote:
>>> Just to be sure, my next step will be to completely disable those
>>> units
>>> and reboot again, but I'm 99% sure that is still going to show the
>>> 30-
>>> second delay.
>> Well, 30sec is annoying. But I thought that it was the 4 min that
>> getting rid of would
>> be the 1st goal.
> I think it was Frank who had the 4 minute delay. I don't think I said
> that mine was 4 minutes, just that there was a delay and I see a blank
> screen with three dots, but apologies if I gave that impression.
No, my delay was about2 minutes. But disabling and masking that
udev.settle unit took care of that.
>
> I no longer have it, but I think I got that from a systemd-analyse plot
> that was posted.
>
> Oh, well. I've gotten to the point where 30s more of boot time doesn't
> bother me that much.
> What bothered me more was when I would go to a directory on which a
> subdirectory of it had
> an NFS mount point with a NAS whose drives were set to power-save. Even
> though I wasn't
> entering the subdirectory I suppose a look-ahead was happening so my
> terminal session
> freeze until the drives came ready.
>