| 1 |
On Mon, Nov 12, 2012 at 3:40 PM, Keith Dart <keith@×××××××××.biz> wrote: |
| 2 |
> Re |
| 3 |
> 20121111192731.149dc082@×××××××××.biz20121111192731.149dc082@×××××××××.bizCADPrc83TGeegA0UXadMEy=B62xZYOE50bqfw+dxDJ8f=FkPTHA@××××××××××.com20121111185801.58964939@×××××××××.bizCADPrc83+5zUQ+GzCSvyHhZa7BZfMqqagZCosktasOY1sdhCx0Q@××××××××××.com20121111175313.54b9acf1@×××××××××.bizCADPrc80EwExpbdcJX1a+aE0DA=iopqw-QOaozmUBJCs5FGew-g@××××××××××.com20121109171149.1d8a3e18@×××××××××.biz509D9C62.9040909@×××××.com509D8E00.4030208@××××××××.sek7k1hn$ce6$1@×××××××××.org, |
| 4 |
> Canek Peláez Valdés said: |
| 5 |
>> That it's weird. What does "systemd-analyze blame" say after a fresh |
| 6 |
>> boot? |
| 7 |
> |
| 8 |
> |
| 9 |
> Something related to consolekit was taking by far the most time. But I |
| 10 |
> thought I was supposed to remove that. So I make USE include |
| 11 |
> -consolekit, and removed console kit, and did "emerge -uD --newuse |
| 12 |
> world", but yet, pambase insists on pulling in consolekit no matter |
| 13 |
> what. I compile it separate (emerge -1 ...) without consolekit USE, and |
| 14 |
> the works. But whenever I update consolekit wants get added again. |
| 15 |
|
| 16 |
Do an "equery depends pambase" and see what it's trying to pull |
| 17 |
pambase with consolekit. Perhaps do you have something in |
| 18 |
/etc/portage/package.use? |
| 19 |
|
| 20 |
> Anyway, right now I have it removed and it does come up faster without |
| 21 |
> it. |
| 22 |
|
| 23 |
Cool. |
| 24 |
|
| 25 |
> But another thing I miss is knowing where it is in the boot process. |
| 26 |
> The systemd doesn't tell me anything. |
| 27 |
|
| 28 |
Actually it does; if you are not using plymouth (or another splash), |
| 29 |
the standard listing of services gets printed on the console, with |
| 30 |
even the "[ OK ]" that a lot of people like. A copy of this output |
| 31 |
gets logged in /var/log/boot.log. |
| 32 |
|
| 33 |
Also, you can boot with systemd.log_target=kmsg |
| 34 |
systemd.log_level=debug, and everything (which is a lot) will get |
| 35 |
stored in the kernel ring buffer, so you can check it out afterwards |
| 36 |
with dmesg. |
| 37 |
|
| 38 |
The thing with this output is that, when properly configured, systemd |
| 39 |
is just too fast for it to be useful; my laptop boots so fast that I |
| 40 |
cannot even see it even if I disable plymouth. |
| 41 |
|
| 42 |
Regards. |
| 43 |
-- |
| 44 |
Canek Peláez Valdés |
| 45 |
Posgrado en Ciencia e Ingeniería de la Computación |
| 46 |
Universidad Nacional Autónoma de México |
Archives