Systemctl - Failed at step Group spawning - raspberry-pi3

I have a problem with automatic running of service. I would like to run a script after user login. Once I have succeeded, however when I tried to automatize it, I'm getting an error and I can't find a mistake.
Service:
Description=Service desc
[Service]
Type=oneshot
Environment=DISPLAY=:0.0
Environment=XAUTHORITY=/home/albert/.Xauthority
Environment=HOME=/home/albert/
User=albert
Group=albert
RemainAfterExit=true
StandardOutput=journal
ExecStart=/home/albert/.startup-script.sh
[Install]
WantedBy=default.target
location:
/home/albert/.config/systemd/user/startup-service.service
permissions: -rw-r--r-- 1 albert albert 306 Jan 2 14:49 startup-service.service
Script
Permissions: -rwxr-x-r-x 1 albert albert 86 Jan 2 13:30 /home/albert/.startup-script.sh
Log
Jan 02 17:03:27 irpi-epsilon systemd[734]: Service desc
-- Subject: Unit UNIT has begun start-up
-- Defined-By: systemd
-- Support: https://www.debian.org/support
--
-- Unit UNIT has begun starting up.
Jan 02 17:03:27 irpi-epsilon systemd[1025]: startup-script.service: Failed at step GROUP spawning /home/albert/.startup-script.sh: Operation not permitted
-- Subject: Process /home/albert/.startup-script.sh could not be executed
-- Defined-By: systemd
-- Support: https://www.debian.org/support
--
-- The process /home/albert/.startup-script.sh could not be executed and failed.
--
-- The error number returned by this process is 1.
Jan 02 17:03:27 irpi-epsilon systemd[734]: startup-script.service: Main process exited, code=exited, status=216/GROUP
Jan 02 17:03:27 irpi-epsilon systemd[734]: Service desc
-- Subject: Unit UNIT has failed
-- Defined-By: systemd
-- Support: https://www.debian.org/support
--
-- Unit UNIT has failed.
--
-- The result is failed.
Jan 02 17:03:27 irpi-epsilon systemd[734]: startup-script.service: Unit entered failed state.
Jan 02 17:03:27 irpi-epsilon systemd[734]: startup-script.service: Failed with result 'exit-code'.
I have called:
systemctl --user daemon-reload
systemctl --user enable startup-script.service
systemctl --user start startup-script
And I'm trying to do it on Raspberry Pi 3.
Thanks a lot for pointing me at any direction.

If somebody has similar problem then just remove
User and Group from service file. I don't know why, but without them it works pretty well. So the final working service description is:
Description=Service desc
[Service]
Type=oneshot
Environment=DISPLAY=:0.0
Environment=XAUTHORITY=/home/albert/.Xauthority
Environment=HOME=/home/albert/
RemainAfterExit=true
StandardOutput=journal
ExecStart=/home/albert/.startup-script.sh
[Install]
WantedBy=default.target

Related

cannot start the squid server

When you make start of squid proxy server:
this massage appears (Job for squid.service failed because the control process exited with error code. See "systemctl status squid.service" and "journalctl -xe" for details.)
How to solve it?
the output:
-- Logs begin at Mon 2022-10-10 14:56:31 +03. –
Oct 13 19:36:40 toj-mgt-uv-fp01..com.sa systemd[1]: Starting Squid Web Proxy Server...
-- Subject: A start job for unit squid.service has begun execution
-- Defined-By: systemd
-- Support: http://www.ubuntu.com/support
-- A start job for unit squid.service has begun execution.
-- The job identifier is 62883.
Oct 13 19:36:40 toj-mgt-uv-fp01..com.sa squid[1051704]: 2022/10/13 19:36:40| ACL not found: fac
Oct 13 19:36:40 toj-mgt-uv-fp01..com.sa squid[1051704]: 2022/10/13 19:36:40| FATAL: Bungled /etc/squid/squid.conf line 6174: http_access allow fac
Oct 13 19:36:40 toj-mgt-uv-fp01..com.sa squid[1051704]: 2022/10/13 19:36:40| Squid Cache (Version 4.10): Terminated abnormally.
Oct 13 19:36:40 toj-mgt-uv-fp01..com.sa squid[1051704]: CPU Usage: 0.010 seconds = 0.000 user + 0.010 sys
Oct 13 19:36:40 toj-mgt-uv-fp01..com.sa squid[1051704]: Maximum Resident Size: 57888 KB
Oct 13 19:36:40 toj-mgt-uv-fp01..com.sa squid[1051704]: Page faults with physical i/o: 0
Oct 13 19:36:40 toj-mgt-uv-fp01..com.sa squid[1051704]: FATAL: Bungled /etc/squid/squid.conf line 6174: http_access allow fac
Oct 13 19:36:40 toj-mgt-uv-fp01..com.sa systemd[1]: squid.service: Control process exited, code=exited, status=1/FAILURE
-- Subject: Unit process exited
-- Defined-By: systemd
-- Support: http://www.ubuntu.com/support
-- An ExecStartPre= process belonging to unit squid.service has exited.
-- The process' exit code is 'exited' and its exit status is 1.
Oct 13 19:36:40 toj-mgt-uv-fp01..com.sa systemd[1]: squid.service: Failed with result 'exit-code'.
-- Subject: Unit failed
-- Defined-By: systemd
-- Support: http://www.ubuntu.com/support
-- The unit squid.service has entered the 'failed' state with result 'exit-code'.
Oct 13 19:36:40 toj-mgt-uv-fp01.****.com.sa systemd[1]: Failed to start Squid Web Proxy Server.
-- Subject: A start job for unit squid.service has failed
-- Defined-By: systemd
-- Support: http://www.ubuntu.com/support
-- A start job for unit squid.service has finished with a failure.
-- The job identifier is 62883 and the job result is failed.
thanks and regards.

speech-dispatcher won't start on manjaro

$ sudo systemctl start speech-dispatcherd.service
Job for speech-dispatcherd.service failed because the control process exited with error code.
See "systemctl status speech-dispatcherd.service" and "journalctl -xeu speech-dispatcherd.service" for details.
ov 10 21:42:32 wiley speech-dispatcher[3128]: [Wed Nov 10 21:42:32 2021 : 132158] speechd: Speech Dispatcher 0.10.2 starting
Nov 10 21:42:32 wiley systemd[1]: speech-dispatcherd.service: Control process exited, code=exited, status=1/FAILURE
░░ Subject: Unit process exited
░░ Defined-By: systemd
░░ Support: https://forum.manjaro.org/c/support
░░
░░ An ExecStart= process belonging to unit speech-dispatcherd.service has exited.
░░
░░ The process' exit code is 'exited' and its exit status is 1.
Nov 10 21:42:32 wiley systemd[1]: speech-dispatcherd.service: Failed with result 'exit-code'.
░░ Subject: Unit failed
░░ Defined-By: systemd
░░ Support: https://forum.manjaro.org/c/support
░░
░░ The unit speech-dispatcherd.service has entered the 'failed' state with result 'exit-code'.
Nov 10 21:42:32 wiley systemd[1]: Failed to start Speech-Dispatcher an high-level device independent layer for speech synthesis..
░░ Subject: A start job for unit speech-dispatcherd.service has failed
░░ Defined-By: systemd
░░ Support: https://forum.manjaro.org/c/support
░░
░░ A start job for unit speech-dispatcherd.service has finished with a failure.
░░
░░ The job identifier is 1578 and the job result is failed.
I'm trying to get speech api to work in firefox. If I type the command to speak from the cli it works fine, but voices array is empty in firefox.
I think speech-dispatcher needs to be running but it won't start on manjaro

Problem running Elasticsearch for Magento project on Ubuntu

I'm trying set up Elasticsearch for Magento 2.4.2 project. I have installed Elasticsearch 7.9.3 and openjdk 11.0.10.
I got the error:
sudo systemctl start elasticsearch:
*Job for elasticsearch.service failed because the control process exited with error code.
See "systemctl status elasticsearch.service" and "journalctl -xe" for details.*
In the /etc/elasticsearch/elasticsearch.yml network settings are:
**network.host:** 127.0.0.1
**http.port:** 9200
journalctl -xe command result:
--
-- An ExecStart= process belonging to unit elasticsearch.service has exited.
--
-- The process' exit code is 'exited' and its exit status is 1.
Mar 10 16:10:46 -ThinkPad-P15s-Gen-1 systemd[1]: elasticsearch.service: Failed with result 'exit-code'.
-- Subject: Unit failed
-- Defined-By: systemd
-- Support: http://www.ubuntu.com/support
--
-- The unit elasticsearch.service has entered the 'failed' state with result 'exit-code'.
Mar 10 16:10:46 -ThinkPad-P15s-Gen-1 systemd[1]: Failed to start Elasticsearch.
-- Subject: A start job for unit elasticsearch.service has failed
-- Defined-By: systemd
-- Support: http://www.ubuntu.com/support
--
-- A start job for unit elasticsearch.service has finished with a failure.
--
-- The job identifier is 10277 and the job result is failed.
Mar 10 16:10:46 -ThinkPad-P15s-Gen-1 sudo[21081]: pam_unix(sudo:session): session closed for user root
Mar 10 16:15:06 -ThinkPad-P15s-Gen-1 sudo[21568]: : TTY=pts/0 ; PWD=/home/; USER=root ; COMMAND=/usr/bin/nano /etc/elasticsearch/>
Mar 10 16:15:06 -ThinkPad-P15s-Gen-1 sudo[21568]: pam_unix(sudo:session): session opened for user root by (uid=0)
Mar 10 16:15:25 -ThinkPad-P15s-Gen-1 sudo[21568]: pam_unix(sudo:session): session closed for user root
Mar 10 16:15:52 -ThinkPad-P15s-Gen-1 wpa_supplicant[853]: wlp0s20f3: WPA: Group rekeying completed with ac:cf:85:db:37:ce [GTK=CCMP]
Looked through all the articles, but none of them helps to fix the situation. Сan someone suggest an idea for a solution to this problem ?
The error data you shared is just identifying the error as 'exited'. This brings me to the guess work to help you with the issue.
From the Elasticsearch.yaml configuration, the port on which the Elasticsearch exposes its API is on 9200. One possibility is that the port might be pre-occupied. Can you look at the port status on your system and confirm it.
Elasticsearch.service file has configurations for memory and CPU. May be the configured memory is not sufficient for it to start.
Hope this helps. For detailed and straight forward instructions on how to install the Magento 2.4.2 this blog written by me might help you.

What is wrong with my systemd service

I have written a golang RestAPI based on labstack/echo and vuejs and have a working version compiled and everything runs nicely when I start it. So far so good.
However when trying to integrate it with systemd to start the process at boot I am stuck. I have a service file.
[Unit]
Description=Server Software Manager
After=network.target
[Service]
Type=simple
ExecStart=/var/gameserver/steam/sman
KillMode=process
User=steam
Group=steam
Restart=on-failure
SuccessExitStatus=2
[Install]
WantedBy=multi-user.target
Alias=sman.service
But everytime I want to start the service I get the following error.
Feb 25 14:17:49 <SERVERNAME> systemd[1]: Stopped Server Software Manager.
Feb 25 14:17:49 <SERVERNAME> systemd[1]: Started Server Software Manager.
Feb 25 14:17:49 <SERVERNAME> systemd[1]: sman.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
Feb 25 14:17:49 <SERVERNAME> systemd[1]: sman.service: Unit entered failed state.
Feb 25 14:17:49 <SERVERNAME> systemd[1]: sman.service: Failed with result 'exit-code'.
Feb 25 14:17:50 <SERVERNAME> systemd[1]: sman.service: Service hold-off time over, scheduling restart.
Feb 25 14:17:50 <SERVERNAME> systemd[1]: Stopped Server Software Manager.
Feb 25 14:17:50 <SERVERNAME> systemd[1]: sman.service: Start request repeated too quickly.
Feb 25 14:17:50 <SERVERNAME> systemd[1]: Failed to start Server Software Manager.
Feb 25 14:19:59 <SERVERNAME> systemd[1]: Started Server Software Manager.
According to google that error is when the Service exits with error code but when I run the Service manually as the steam user it does not do that.
My assumption is that something is wrong with that unit file but I don't know what. And Systemd-analyze has also not complained.
I am completely lost and thankful for any leads you might have help debug this.
The output of jounarlctl -xfe -u sman:
Feb 26 14:18:23 <SERVERNAME> systemd[1]: Started Server Software Manager.
-- Subject: Unit sman.service has finished start-up
-- Defined-By: systemd
-- Support: http://lists.freedesktop.org/mailman/listinfo/systemd-devel
--
-- Unit sman.service has finished starting up.
--
-- The start-up result is done.
Notes:
OS: Ubuntu 16.04 LTS
I finally managed to find out what was wrong. Systemd needs all Env Variables explicitly mentioned. However it looks like I had loads of troubles reloading systemd services. This is what i've ended up doing. the service needs to be stoped, the .service file moved to a location systemd does not read, after that reload systemd with systemctl daemon-reload and wait until systemctl status $SERVICE claims there is no such service. (Systemd might still reread the cache if status claims anything else) After that add the .service file again to the appropriate location and voila it finaly worked.

Kubernetes Installation with Vagrant & CoreOS behind proxy

I am behind a proxy server and following the "Kubernetes Installation with Vagrant & CoreOS" steps listed here: https://coreos.com/kubernetes/docs/latest/kubernetes-on-vagrant.html
After finalizing the install, when I run
$ kubectl get nodes
I get the error.
Unable to connect to the server: Service Unavailable
e1, c1, and w1 are up and I can issue $vagrant ssh each of them.
when I check the w1, I have seen docker service was not running with the error listed below.
----------------------------------------------------------------------------
-- Unit docker.service has failed.
--
-- The result is dependency.
Aug 19 04:09:25 w1 systemd[1]: docker.service: Job docker.service/start failed with result 'dependency'.
Aug 19 04:09:25 w1 systemd[1]: flanneld.service: Unit entered failed state.
Aug 19 04:09:25 w1 systemd[1]: flanneld.service: Failed with result 'exit-code'.
Aug 19 04:09:30 w1 systemd[1]: flanneld.service: Service hold-off time over, scheduling restart.
Aug 19 04:09:30 w1 systemd[1]: Stopped Network fabric for containers.
-- Subject: Unit flanneld.service has finished shutting down
-- Defined-By: systemd
-- Support: http://lists.freedesktop.org/mailman/listinfo/systemd-devel
--
-- Unit flanneld.service has finished shutting down.
Aug 19 04:09:30 w1 systemd[1]: Starting Network fabric for containers...
-- Subject: Unit flanneld.service has begun start-up
-- Defined-By: systemd
-- Support: http://lists.freedesktop.org/mailman/listinfo/systemd-devel
--
-- Unit flanneld.service has begun starting up.
Aug 19 04:09:30 w1 rkt[6888]: image: using image from file /usr/lib/rkt/stage1-images/stage1-fly.aci
Aug 19 04:09:31 w1 rkt[6888]: image: searching for app image quay.io/coreos/flannel
Aug 19 04:09:31 w1 rkt[6888]: run: discovery failed
Aug 19 04:09:31 w1 systemd[1]: flanneld.service: Main process exited, code=exited, status=1/FAILURE
Aug 19 04:09:31 w1 systemd[1]: Failed to start Network fabric for containers.
-- Subject: Unit flanneld.service has failed
-- Defined-By: systemd
-- Support: http://lists.freedesktop.org/mailman/listinfo/systemd-devel
--
-- Unit flanneld.service has failed.
--
-- The result is failed.
Aug 19 04:09:31 w1 systemd[1]: flanneld.service: Unit entered failed state.
Aug 19 04:09:31 w1 systemd[1]: flanneld.service: Failed with result 'exit-code'.
----------------------------------------------------------------------------
I am guessing that problem is because of I am behind the proxy. Before running the install steps I issue the commands
$export "HTTP_PROXY=http://http-proxy.xxxxxx.com:8080"
$export "HTTPS_PROXY=http://http-proxy.xxxxxx.com:8080"
$export "http_proxy=http://http-proxy.xxxxxx.com:8080"
$export "https_proxy=http://http-proxy.xxxxxx.com:8080"
Do you know if this is enough for the installation behind proxy or do I need to add proxy settings to somewhere else.
Thank you in advance,
turgos
The variables you're exporting are valid only in your current shell session, they are not available to your flannel systemd unit.
Create the following drop-in inside the systemd unit directory, and then reload the daemon with systemctl daemon-reload, it should fix your issue with flannel:
/etc/systemd/system/flannel.service.d/proxy.conf:
[Service]
Environment="HTTP_PROXY=http://http-proxy.xxx:8080"
Environment="...
A similar example is available in the CoreOS documentation: Customizing Docker

Resources