[SOLVED] Day way to get systemd to show deps for particular target?

[charlweed]

I read the manual for systemD and systemctl, and I could not see a way to display the dependencies of local-fs.target. Is there one?

My system hangs at boot, somewhere before the local-fs.target completes.
Not only do I know know what is failing, I don't know what it is trying to do.

I can boot to the emergency target. But that does not trigger the problem. I have the logging options on, but it FREEZES when I try the local-fs.target, so I never get to open the logs.

I'd like to know what it is trying to do, so I can step through manually.

I HATE HATE HATE!! systemD. What a horrible DISASTER of a system.

---------- Post added at 05:55 PM ---------- Previous post was at 05:08 PM ----------

Even more fun.
I boot to emergency.
Manually start each local-fs.target.wants/* then local-fs.target
It waits a long tome for something, then it "works" but does not mount my lvm filesystems.
I try again for default.target, and this time local-fs.target fails with a dependency, and the system hangs again.
syslog is not available at the time of the hang.
What ever is the first fail, is lost when the console scrolls.
I's looking a lot like I must do a full reinstall of this system.

[DBelton]

You may want to try letting it sit there for awhile when it "hangs"

Currently, if a filesystem check is forced at boot time, systemd does not put a progress indicator on the screen, and it seems like the system is hung, when in fact it is checking a filesystem.

This may be the case here since you say it's hanging about the time local-fs is running.

Just let it sit for awhile and see what happens.

[cazo]

To answer your question:


systemctl --no-pager show -p "Wants" local-fs.target

Wants=media.mount fedora-storage-init-late.service var-lock.mount fsck-root.service remount-rootfs.service fedora-storage-init.service systemd-remount-api-vfs.service fedora-readonly.service var-run.mount


These are in /lib/systemd/system/local-fs.wants/.

This is (adapted) from the systemd Tips & Tricks page

I believe DBelton is correct about it probably being an 'fsck' pause (I wish plymouth would show some type of indicators). You could try either hitting <ESC> (that usually switches back and forth between plymouth & a console screen) or hitting <Tab> during grub's boot screen, and editing the boot line to remove (temporarily) rhgb (which will show you the 'old style' boot screen, with various messages).

[charlweed]

Thanks guys,
systemctl --no-pager show -p "Wants" local-fs.target
is the command I wanted. but it did not help me find the ultimate cause of the dependency failure.

I THINK the problem is in cryptsetup, but I will keep testing.
My "upgrade" to 2.6.40-4.fc15.x86_64 failed, and now I get :

 
modprobe: FATAL: Error inserting padlock_sha (/lib/modules/2.6.40-4.fc15.x86_64/kernel/drivers/crypto/padlock-sha.ko): No such device
  padlock_sha: VIA PadLock Hash Engine not detected.

Those may be a red herring though.
Time for new thread.

[DBelton]

From what I have seen on that error...

It is a warning that is telling you that the hardware doesn't exist, which would make a lot of sense. The VIA Padlock Security Engine is only on VIA x86 processors. Do you have a VIA processor in your machine?