Mailinglist Archive: opensuse (626 mails)

< Previous Next >
Re: [opensuse] Login weirdness
On 06/11/2018 14.12, Michael Fischer wrote:
On Tue, Nov 06, Carlos E. R. wrote:
On 05/11/2018 22.13, Michael Fischer wrote:
On Mon, Nov 05, Carlos E. R. wrote:

And you have problems login in there, for a while, then it works? Very
strange.

And you have already replaced the /home disk? The old disk is
disconnected? I mean, the data cable removed?

Nope. At the point where I've got all the equipment I've ordered and
install that new SSD for /home, I'm going to install Leap 15. Currently
all the same stuff, still connected.

Play with "systemd-analyze". Tab tab for options. Start with "blame".
Then "critical-chain". You may see what service, if any, is taking long
to start.

In Critical chain the '+' indicates how long a service takes to complete
start; and when this value does appear, it is that it takes long. Those
lines may show in red.

The multiuser target is started early, but many other services are
started later and take long to complete.

I append `blame` - though I don't see anything striking. I'm used to
wicked taking a while (relatively speaking), but login seems to take
very little time.

Try with critical chain when you can, please.

blame only shows if a service is too slow, but not if it has impact.

cer@Telcontar:~> systemd-analyze critical-chain
The time after the unit is active or started is printed after the "@"
character.
The time the unit takes to start is printed after the "+" character.

cer@Telcontar:~> systemd-analyze blame
3min 41.381s purge-kernels.service
37.341s systemd-cryptsetup@cr_cripta.service
3min 41.381s purge-kernels.service
37.341s systemd-cryptsetup@cr_cripta.service
16.353s clamd.service
7.834s spamd.service
6.180s rpmconfigcheck.service
5.796s fstrim.service
5.034s wicked.service
4.374s SuSEfirewall2.service
1.988s home_aux.mount
1.985s home1.mount
1.609s data-vmware.mount
1.357s ntpd.service



multi-user.target @1min 4.525s
└─NetworkManager-dispatcher.service @1min 19.687s +18ms
└─basic.target @40.576s
└─sockets.target @40.565s
└─avahi-daemon.socket @40.554s
└─sysinit.target @40.457s
└─sys-fs-fuse-connections.mount @1min 35.814s +13ms
└─systemd-modules-load.service @382ms +1.221s
└─systemd-journald.socket
└─-.mount
└─system.slice
└─-.slice

(and the system is on ssd...)



The fsck's are on the two partitions on the rust disk. Is there some
dependency between login(1) and /home being "available"?

Maybe... Dunno. login of a user should fail, but for a reason that says
that the disk is not ready.



5.323s wicked.service
1.096s spamd.service
321ms vboxdrv.service
245ms systemd-journal-flush.service



--
Cheers / Saludos,

Carlos E. R.
(from 42.3 x86_64 "Malachite" at Telcontar)

< Previous Next >
Follow Ups