[SOLVED] Re: Very slow boot in F27 after recent update
by Gianluca Cecchi
On Fri, May 4, 2018 at 9:32 AM, Gianluca Cecchi <gianluca.cecchi(a)gmail.com>
wrote:
> I'm not subscribed to receive all the e-mails, so I answer to my own.
> Hope not to scramble the thread order...
>
> Thanks for answering.
> Here the output requested.
> Possibly before the update, chronyd and gssproxy (and I see also libvirtd)
> went into a sort of parallel background not preventing display manager
> login completion?
> In fact in about 10-13 seconds (I have an SSD disk on the laptop) I had
> the gdm login page....
>
> [g.cecchi@ope46 ~]$ sudo systemd-analyze time
> Startup finished in 1.379s (kernel) + 1.965s (initrd) + 1min 49.668s
> (userspace) = 1min 53.012s
> [g.cecchi@ope46 ~]$
>
> [g.cecchi@ope46 ~]$ sudo systemd-analyze blame
> 1min 30.134s gssproxy.service
> 1min 15.937s chronyd.service
> 30.043s NetworkManager-wait-online.service
> 17.876s libvirtd.service
> 1.102s dracut-initqueue.service
> 1.015s lvm2-monitor.service
>
>
It seems latest updates have solved my problems, both if network cable is
connected and without it.
My network card on laptop is this one from "lspci -v":
05:00.0 Ethernet controller: Qualcomm Atheros AR8151 v2.0 Gigabit Ethernet
(rev c0)
Subsystem: ASUSTeK Computer Inc. Device 1851
Flags: bus master, fast devsel, latency 0, IRQ 35
Memory at dc800000 (64-bit, non-prefetchable) [size=256K]
I/O ports at 9000 [size=128]
Capabilities: [40] Power Management version 3
Capabilities: [48] MSI: Enable+ Count=1/1 Maskable- 64bit+
Capabilities: [58] Express Endpoint, MSI 00
Capabilities: [6c] Vital Product Data
Capabilities: [100] Advanced Error Reporting
Capabilities: [180] Device Serial Number ff-bd-21-9b-14-da-e9-ff
Kernel driver in use: atl1c
Kernel modules: atl1c
It seems combination of latest kernel (4.16.6-202.fc27.x86_64) and latest
packages I have again my old behavior.
With network cable connected to an automatic dhcp connection:
[g.cecchi@ope46 ~]$ sudo systemd-analyze time
Startup finished in 1.386s (kernel) + 1.328s (initrd) + 8.891s (userspace)
= 11.606s
[g.cecchi@ope46 ~]$
[g.cecchi@ope46 ~]$ sudo systemd-analyze blame
4.568s ntpdate.service
2.145s NetworkManager-wait-online.service
1.168s cups.service
1.158s plymouth-quit-wait.service
1.019s abrtd.service
819ms lvm2-monitor.service
716ms systemd-udev-settle.service
712ms libvirtd.service
. . .
Without network cable connected (and without wirelsss autoconnect)
[g.cecchi@ope46 ~]$ sudo systemd-analyze time
Startup finished in 1.389s (kernel) + 2.002s (initrd) + 8.702s (userspace)
= 12.094s
[g.cecchi@ope46 ~]$
and
[g.cecchi@ope46 ~]$ sudo systemd-analyze blame
6.337s NetworkManager-wait-online.service
1.189s cups.service
1.181s dracut-initqueue.service
1.021s abrtd.service
971ms lvm2-monitor.service
886ms plymouth-quit-wait.service
865ms systemd-udev-settle.service
682ms firewalld.service
641ms libvirtd.service
. . .
dnf summary for latest updates has been:
Installed:
kernel.x86_64 4.16.6-202.fc27 kernel-core.x86_64
4.16.6-202.fc27
kernel-devel.x86_64 4.16.6-202.fc27 kernel-modules.x86_64
4.16.6-202.fc27
kernel-modules-extra.x86_64 4.16.6-202.fc27
python2-ntlm-auth.noarch 1.1.0-1.fc27
Upgraded:
ansible.noarch 2.5.2-1.fc27
cinnamon-themes.noarch 1:1.6.2-1.fc27
copy-jdk-configs.noarch 3.3-9.fc27
httpd.x86_64 2.4.33-5.fc27
httpd-filesystem.noarch 2.4.33-5.fc27
httpd-tools.x86_64 2.4.33-5.fc27
hwdata.noarch 0.312-1.fc27
kdiagram.x86_64 2.6.1-2.fc27
kernel-headers.x86_64 4.16.6-202.fc27
libpsl.x86_64 0.18.0-2.fc27
mate-panel.x86_64 1.20.1-5.fc27
mate-panel-libs.x86_64 1.20.1-5.fc27
mint-themes.noarch 1:1.6.2-1.fc27
mint-themes-gtk3.noarch 1:1.6.2-1.fc27
mint-x-icons.noarch 1.4.7-1.fc27
mint-y-icons.noarch 1.1.4-1.fc27
mint-y-theme.noarch 1:1.6.2-1.fc27
mod_ssl.x86_64 1:2.4.33-5.fc27
opencv.x86_64 3.2.0-15.fc27
opencv-contrib.x86_64 3.2.0-15.fc27
opencv-core.x86_64 3.2.0-15.fc27
powerdevil.x86_64 5.12.4-2.fc27
publicsuffix-list-dafsa.noarch 20180419-1.fc27
python2-opencv.x86_64 3.2.0-15.fc27
python2-requests_ntlm.noarch 1.1.0-1.fc27
python2-winrm.noarch 0.3.0-1.fc27
selinux-policy.noarch 3.13.1-283.34.fc27
selinux-policy-devel.noarch 3.13.1-283.34.fc27
selinux-policy-targeted.noarch 3.13.1-283.34.fc27
spirv-tools-libs.i686 2018.3.0-0.1.20180407.git26a698c.fc27
spirv-tools-libs.x86_64 2018.3.0-0.1.20180407.git26a698c.fc27
vim-common.x86_64 2:8.0.1788-1.fc27
vim-enhanced.x86_64 2:8.0.1788-1.fc27
vim-filesystem.noarch 2:8.0.1788-1.fc27
vim-minimal.x86_64 2:8.0.1788-1.fc27
vulkan.i686 1.1.73.0-1.fc27
vulkan.x86_64 1.1.73.0-1.fc27
vulkan-filesystem.noarch 1.1.73.0-1.fc27
Cheers,
Gianluca
4 years
cpupower
by JD
Manpage of cpu power make no mention of the
selection of governor.
How can set the governor to be the user mode governor
instead of the performance governor and not the ondemand governor?
If I can permanently set that governor, next I want to
set the cpu frequency (for all cores) to 2134000Hz permanently.
Reason I am posting this is because the normal speed of the cores
is 2.8GHz, and that is causing numerous kerneloops interrupts (overheating).
Fans are at full speed all the time, as I can hear them :) :)
When I set the speed manually (as root) like this:
for i in 0 1 2 3 4 5 6 7 8; do
echo 2134000 > cpu$i/cpufreq/scaling_max_freq
done
then kerneloops are greatly reduced.
Every kerneloops is accompanied by mce:
So, it has to be the heat.
But I do not want to do this loop every time I boot up.
I want to have set once and for all!!!
Thanx!!!
4 years
>60gb of reserved memory??
by Susi Lehtola
Dear all,
I just noticed a very weird problem on my workstation. Although the
machine has 64 GB of RAM installed,
$ head -n 1 /proc/meminfo
MemTotal: 2400748 kB
that is, only a few gigabytes are actually available.
Unfortunately I don't have physical access to the machine, which is a HP
desktop bought last fall. According to dmesg, the model is something like
[ 0.000000] DMI: HP HP EliteDesk 800 G3 TWR/8298, BIOS P01 Ver. 02.06
06/09/2017
Also according to dmesg, it appears that the kernel *does* see all 64
gigs, BUT OVER 60 GB IS RESERVED MEMORY!!
[ 0.000000] Memory: 2326720K/66985176K available (12300K kernel code,
1568K rwdata, 3836K rodata, 2124K init, 1368K bss, 64658456K reserved,
0K cma-reserved)
Does anyone have any idea of what could be going on?
I am running an up-to-date Fedora 27 x86_64 with kernel
4.16.15-200.fc27.x86_64.--
Susi Lehtola
Fedora Project Contributor
jussilehtola(a)fedoraproject.org
4 years
Could LightDM be bad?? (LONG)
by Beartooth
I keep three PCs on my desk, running Fedora behind a KVM switch: #1 is
my present biggest fastest, #2 its predecessor, and #3 that one's
predecessor; I try try keep them as similar as is feasible.
They've all been having troubles for months, which have only gotten worse.
Under F27, they kept filling up with some sort of cruft, to the point
of refusing dnf upgrade; but when I found any of the cruft, it was in
places where I dared not lay about me with a cyber-battleaxe. I jumped
to F28 the day of release.
At this point, none of the three is usable at all without some dodge or
other.
#2 and #3 have each been wiped, twice, with different releases of DBAN,
followed each time with a netinstall of F28 with Mate, and then dnf
upgrade daily. All their boot messages show green OK. Those end with
"Started Light Display Manager."
Usually they proceed to a login box in front of the F28 display of blue
light-conducting fibers. That accepts my password, churns a little way,
and ends with a monitor full of blue horizontal lines, plus a few white
lines in the middle and at the bottom; there is nothing legible.
By doing Alt-Ctrl-F2, I get a display which lets me log in as root; I
keep doing dnf upgrades, and trying startx, every day. I also reboot
whenever there's a kernel change.
Startx always flashes a couple times, then fails with a few lines of
text, beginning with an attempt to adopt my monitor size (called "1920
1080" -- not 1920x1080),losing contact, and ending by saying that the
Xserver "terminated successfully."
I'd like to try replacing LDM with something else; how do I do that??
(I like Mate well, and would prefer not to swap it out,too)
Finally, #1 has two F28 kernels -- which do no better than on #2 nor #3
-- and a rescue kernel from F26. Sic. Twenty-six, not 27 nor 28. But
that rescue kernel does support almost my whole GUI (I miss Pan badly.),
and I don't have to descend to the console.
#1 got F28 by upgrading with dnf from F27, and still has a lot of stuff
on it that I'd rather copy to #2 and #3 directly than by way of solid
state storage, if I dare risk malware. (I don't know where else all the
cruft could have come from, apart from malware, but I hope someone here
does.)
--
Beartooth Staffwright, Not Quite Clueless Power User
Remember I know little (precious little!) of where up is.
4 years
caja mount of network drives
by Robert McBroom
When I look where caja used to mount network drives I see
/run/user/1001:
/usr/bin/ls: cannot access '/run/user/1001/./doc': Permission denied
/usr/bin/ls: cannot access '/run/user/1001/./gvfs': Permission denied
total used in directory 4 available 358772
drwx------. 9 rm3 rm3 280 Jun 17 23:25 .
drwxr-xr-x. 3 root root 60 Jun 17 01:17 ..
srw-rw-rw-. 1 rm3 rm3 0 Jun 17 01:17 bus
drwx------. 3 rm3 rm3 60 Jun 17 01:17 dbus-1
drwx------. 2 rm3 rm3 60 Jun 17 07:38 dconf
d?????????? ? ? ? ? ? doc
d?????????? ? ? ? ? ? gvfs
srw-------. 1 rm3 rm3 0 Jun 17 07:34 kdeinit5__0
drwx------. 2 rm3 rm3 60 Jun 17 07:35 keyring
srwxrwxr-x. 1 rm3 rm3 0 Jun 17 01:17 klauncherTJ1413.1.slave-socket
srwxrwxr-x. 1 rm3 rm3 0 Jun 17 07:34 klauncherTJ7613.1.slave-socket
-rw-rw-r--. 1 rm3 rm3 67 Jun 17 07:34 KSMserver__0
drwx------. 2 rm3 rm3 80 Jun 17 01:18 pulse
drwxr-xr-x. 3 rm3 rm3 100 Jun 17 01:17 systemd
Robert
4 years
Redshift has an issue with geoclue2
by ToddAndMargo
Hi All,
Fedora 28, x64
redshift-gtk-1.11-8.fc28.x86_64
redshift-1.11-8.fc28.x86_64
geoclue2-libs-2.4.10-2.fc28.x86_64
geoclue2-2.4.10-2.fc28.x86_64
When starting Redshift, I get the following error
$ redshift
Trying location provider `geoclue2'...
Using provider `geoclue2'.
Unable to start GeoClue client:
GDBus.Error:org.freedesktop.DBus.Error.NoReply: Message
recipient disconnected from message bus without replying.
Unable to connect to GeoClue.
Unable to get location from provider.
the pop up says:
GDBus.Error" org.freedesktop.DBus.Error.AccessDenied
Any word of wisdom? It worked under Fedora 27.
-T
4 years
Fedora's OwnCloud and NextCloud Broke - What I did
by Clifford Snow
When I upgraded to F28, OwnCloud stopped working. Another reported the same
problem with NextCloud. The problem has been reported in bugzilla [1] as
Cannot login after upgrading from F27 to F28. Patrick Reeb provided a fix
with a fix-owncloud [2] script which downgrades php-sabre to F25.
Because both of the Fedora packages of OwnCloud and NextCloud are dated I
decided to use the NextCloud zip package of their current version 13.
Basically what I did was to install Fedora's NextCloud package to get the
basic configuration files and then removed the package so all what was left
was configuration files with a .rpmsave extension. Using the
nextcloud-13.0.4.zip package [3] I moved the files to my nextcloud
directory from the Fedora's package.
After a few tweaks to /etc/httpd/conf/httpd.conf files, the login prompt
appeared on my localhost/nextcloud. I was unable to configure my
postgresql database from the web browser, so I used the command line tool
[4] (occ) to finish the install.
I did try the snap install of the nextcloud-server, but couldn't get it to
work. However, the snap nextcloud-client works like a charm.
[1] https://bugzilla.redhat.com/show_bug.cgi?id=1562949
[2] https://bugzilla.redhat.com/attachment.cgi?id=1431595
[3] https://nextcloud.com/install/#instructions-server
[4] https://nextcloud.com/install/#instructions-server
--
@osm_seattle
osm_seattle.snowandsnow.us
OpenStreetMap: Maps with a human touch
4 years