Re: Problem source verified -Re: Jessie (8.0) slow to boot
On Fri 02 Sep 2016 at 17:44:39 (+0200), Nicolas George wrote:
> Le septidi 17 fructidor, an CCXXIV, David Wright a écrit :
> > Basically, if "KiB Swap: ... 0 used," is anything but zero, I might as
> > well reboot; any idea of getting work done while swapping is risible,
>
> If you have to look at the output of a command do notice the swap is in use,
> that means you were able to getting work done very well.
Clever, but clearly not correct.
If the computer is nearby, as is typical, then the first two things
you notice are the noise from the disk and the slowdown.
So then the question is whether the computer is just generally
overloaded or, more usually, the browser is out of control.
The latter can be no more than the browser consuming more than
100% CPU in which case simply killing it is enough to restore order.
On the laptop, it might be possible to tell high CPU consumption
(obviously accompanied by disk activity) from swapping by the pitch of
the fan and the increasing discomfort in the right thigh, but $ top
is easier just so long as fvwm is able to switch to a suitable screen,
which is not a given.
There have been kernels when the right thigh has cried "wolf" for
too long, and then the best indication of increasing CPU temperature
is the colour of the root window, which switches to red at 80° and
brown at 85°, though that's not much help if a video is playing
full-screen. (My current record temperature peak is 96.5°.)
Forget all this if I'm doing the work on the computer in the basement¹
or the one in the attic². How am I going to hear disk swapping when I'm
sitting under a ceiling fan with the radio on? More than 80% of the
former's and 100% of the latter's use is remotely from the laptop.
> > But I guess you may just mean that systemd
> > cares about /etc/fstab in a way that sysvinit didn't
>
> sysvinit is just a bunch of shell script. Error codes were lost all the time
> and nobody cared.
Looks like my guess was correct, then.
¹ anti-tornado. ² anti-flash flooding.
Cheers,
David.
Reply to: