Error Message at Shutdown / Reboot

As nobody confirmed this issue

My apologies, I have totally missed this thread. :(

I have to confirm the issue (is it an issue?), exactly as Bequimão describes:

I get at shutdown or reboot frequently, but not always the following message:

systemd-journald [###]: Failed to send streamfile descriptor to servicemanager: Transport endpoint is not connected.

It is difficult to spot. The process-id refers to the same systemd-journald.

Information about my system according to the Sysinfo tool in Octopi: https://gist.github.com/5bfc50e2c9645c3f980648074d351e73

(I hope I got it right, this is the first time that I use this Sysinfo tool.)

And no: I don't use btrfs, but ext4 for both / and /home.

To add a few details:

On my machine this behavior occurs very rarely. I cannot give you exact numbers, but let's say, out of 20/25 shutdowns (I don't reboot my system really often) it occurs once, so the last occurrence might date from the beginning of May - I simply can't remember. That's why I did not consider this behavior a severe issue and did not open a thread about it myself.

When it occurs, I get the message Bequimão quotes several times, approximately 20 lines one below the other, but no longer than a second or so, rather less; it is hard to even read the message(s) before the system finally shuts down.

I'm not noticed it, i think this not really matters. systemd-journald is a demon who writes logs in /var/log/journal/. It's looks like failed to initialize output file to write in. I think this can be happen if you mount /var/log in RAM or some mount restrictions.

i think this not really matters. It's looks like failed to initialize output file to write in. I think this can be happen if you mount /var/log in RAM or some mount restrictions.

I guess you are right.

I agree with Bequimão and bvbfan that this seems to be a Systemd issue.

Concerning the mount points - well, maybe I have to clarify:

$ lsblk

NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT

sda 8:0 0 298,1G 0 disk

├─sda1 8:1 0 29,3G 0 part /

├─sda2 8:2 0 7,7G 0 part [SWAP]

└─sda3 8:3 0 261,1G 0 part /home

sr0 11:0 1 1024M 0 rom

A completely conservative partitioning scheme for a private user's desktop PC, nothing sophisticated: /var/log has no extra partition, and nothing is mounted in RAM. To underline:

$ cat /etc/fstab

[...]

UUID=xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx / ext4 defaults,noatime 0 1

UUID=xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx swap swap defaults,noatime 0 0

UUID=xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx /home ext4 defaults,noatime 0 2

Nevertheless, I get those messages from time to time at shutdown.

Any of you using the repo? Changes for this with systemd 230?

Any of you using the repo?

Sorry, no, still on systemd 229. Using an innovative OS, I've chosen a completely conservative approach. ^^

But another question: Can this issue have something to do with watchdog?

If this idea is stupid, then simply forget about it. If it is, however, possible, then I will expand on this.

(My apologies in advance, concerning PCs and all that geek stuff, I am a complete autodidact, no kernel hacker, nothing.)

@ I activated and did an upgrade, but no systemd 230 shown.

Best regards,

Bequimão

Post your /etc/pacman.conf, sounds like you did not add build correctly (it needs to be the very first repo, above core)

This issue with SystemD gave me the idea that there could be a relation with the issue you reported (those "Failed to send streamfile descriptor to servicemanager" lines) and messages I get quite often at shutdown.

@ Bequimão: So, my question is: Do you sometimes at shutdown get messages similar to

watchdog did not stop

Unlike those "Failed to send streamfile descriptor to servicemanager" lines, I get watchdog related complaints quite often. But just now I do not have an example in my own journal that I could cite. ^^

Maybe, this is a completely wrong direction, then "Sorry" again. And no worry: This will be my last post in this thread. ^^

Though the very first message at boot is still "starting version 229"

You can surpress this message altogether.

To have the correct version showing, make sure to always update the kernel image on systemd updates mkinitcpio -p linux repeat for linux-next, if you use that kernel too.