Re: Headless server just got suspended by updating systemd
On 27/11/14 01:30, Patrick Wiseman wrote:
> On Tue, Nov 25, 2014 at 11:50 PM, Scott Ferguson
> <scott.ferguson.debian.user@gmail.com> wrote:
>> On 26/11/14 12:46, Andrew McGlashan wrote:
>>> On 26/11/2014 11:07 AM, Scott Ferguson wrote:
>>>> And *what do we know about the original "customisation"* that was
>>>> made - which may have 'some' bearing on the "badness" of the
>>>> upgrade/update "defaults"??
>>>
>>> Oh, I get it,
>>
>> Patently, and demonstrably - you do *not*.
>>
>> Assuming your best intentions[*1], let me try and make it simple for you:-
>> It's considered Best Practise to *identify*[*2] the problem *before*
>> attempting to *solve* it.
>
> I identified the problem (my always closed and never before suspended
> laptop, while I was upgrading through an ssh session, suddenly
> suspended during the upgrade); I solved it (found the laptop in
> suspend mode, opened the lid, and the upgrade completed). Next problem
> identified (now when I closed the lid, the laptop - on AC power but
> with no functioning monitor - suspended); I solved it (by Googling and
> finding the solution previously posted).
If only more people followed your example and:-
;patiently answered questions instead of dump and run
;did their research
;published there results
;carefully drew conclusions on causes - or didn't when insufficient data
is available, especially when the context is contentious (kudos to you)
>
> Now, it may be that the "problem" is deeper, and that, as someone else
> has suggested, logind by default takes over a function previously
> performed by pm-utils, using which (with the xfce GUI interface) I had
> told the laptop NOT to suspend even though the lid was closed.
pm-utils? Thanks, I do appreciate how hard it can be sometimes to
recall/find notes of old setups.
Do you recall how you disabled suspend? Perhaps using an xfce app/config?
> (It
> lives that way, functioning as a headless server.)
pm-utils? Thanks, I do appreciate how hard it can be sometimes to
recall/find notes of old setups.
You may have noted my previous comments about not installing
laptop-detect when using a laptop on mains (they make great
firewall/router/gateways). I've also found power-saving problematic when
employing laptops for those purposes and found removing that package the
simplest way to deal with the problems.
>
> [...]
>
>> In this instance - while Patrick has done the "right" thing by finding
>> and publishing a solution, he hasn't provided any information about what
>> he originally used for suspend - and how, or if, the laptop was not
>> previously suspending. Only that it was a long time ago and he doesn't
>> remember
>> (<[🔎] CAJVvKsOHyOymUWrDPA1GLC7daqDeFkcPhj_FDV3SXj3wUS-DGw@mail.gmail.com>).
>
> Not sure what that's supposed to be a link to,
It's the id of your original post. I 'should' have searched the mail
archives to find a reference that is easier for the majority to find.
https://lists.debian.org/debian-user/2014/11/msg01613.html
> but I thought I had
> made it clear that the laptop lid was closed and the laptop was NOT
> suspending. And, as I also said previously, I used the xfce GUI power
> manager to control what happened when the lid closed;
Somehow I missed that. My apologies.
*That* (xfce power manager) is what you 'should' file the bug report
against.
> presumably it
> was talking to something further under the hood (pm-utils?) which has
> now apparently been overridden by logind.
>
> [...]
>
>> So before debating the upgrade process it 'might' be best to try and
>> recreate the scenario don't you think?
>
> That would probably not be possible,
Not necessary given that we now know where the setting was made. It
'should' be possible to determine how xfce power manager disables
suspend on lid close - and which part of the dist-upgrade doesn't
respect that setting.
> although people in this thread
> have made what seem to me to be good guesses about what probably
> happened. But the upgrade was of over a thousand packages. All I was
> trying to do was to provide information which might prove helpful to
> others. But apparently one can have every good intention of not
> starting a fight on debian-user and a fight breaks out anyway.
Sadly yes.
>
> Be nice to one another,
> Patrick
>
>
Kind regards
Reply to: