I see the board spends a lot of time on internet connections. Can I start this service later, after the full load? Thanks.
For deeper analyze you could try to visualize it a bit more
systemd-analyze critical-chain
or as picture
systemd-analyze plot > /tmp/blame.svg
This should give an overview of the critical path
if possible, try to use SSH and copy stuf directly from command line, instead of creating picture.
Can you create the analyze view pls
systemd-analyze plot > /tmp/blame.svg
Thanks for help.
File here: Fastupload.io — Download — blame.svg
I see you are using NetworkManager.service? At least it is installed. Because NetworkManager is conflicting service to ifupdown. You should use just one. Not both. Did you installed HA Supervised?
Can you check the log of the ifup@eth0.service
journalctl -u ifup@eth0.service
I plan to use nmcli for it needed NetworkManager.
journalctl -u ifup@eth0.service:
-- Journal begins at Mon 2022-12-12 19:57:07 GMT, ends at Mon 2022-12-12 21:17:01 GMT. --
Dec 12 19:57:09 DietPi systemd[1]: Starting ifup for eth0...
Dec 12 19:57:10 DietPi dhclient[459]: Internet Systems Consortium DHCP Client 4.4.1
Dec 12 19:57:10 DietPi ifup[459]: Internet Systems Consortium DHCP Client 4.4.1
Dec 12 19:57:10 DietPi ifup[459]: Copyright 2004-2018 Internet Systems Consortium.
Dec 12 19:57:10 DietPi ifup[459]: All rights reserved.
Dec 12 19:57:10 DietPi ifup[459]: For info, please visit https://www.isc.org/software/dhcp/
Dec 12 19:57:10 DietPi dhclient[459]: Copyright 2004-2018 Internet Systems Consortium.
Dec 12 19:57:10 DietPi dhclient[459]: All rights reserved.
Dec 12 19:57:10 DietPi dhclient[459]: For info, please visit https://www.isc.org/software/dhcp/
Dec 12 19:57:10 DietPi dhclient[459]:
Dec 12 19:57:10 DietPi dhclient[459]: Listening on LPF/eth0/f6:a5:ed:fd:ae:23
Dec 12 19:57:10 DietPi dhclient[459]: Sending on LPF/eth0/f6:a5:ed:fd:ae:23
Dec 12 19:57:10 DietPi ifup[459]: Listening on LPF/eth0/f6:a5:ed:fd:ae:23
Dec 12 19:57:10 DietPi ifup[459]: Sending on LPF/eth0/f6:a5:ed:fd:ae:23
Dec 12 19:57:10 DietPi ifup[459]: Sending on Socket/fallback
Dec 12 19:57:10 DietPi ifup[459]: DHCPREQUEST for 192.168.31.238 on eth0 to 255.255.255.255 port 67
Dec 12 19:57:10 DietPi dhclient[459]: Sending on Socket/fallback
Dec 12 19:57:10 DietPi dhclient[459]: DHCPREQUEST for 192.168.31.238 on eth0 to 255.255.255.255 port 67
Dec 12 19:57:14 DietPi dhclient[459]: DHCPREQUEST for 192.168.31.238 on eth0 to 255.255.255.255 port 67
Dec 12 19:57:14 DietPi ifup[459]: DHCPREQUEST for 192.168.31.238 on eth0 to 255.255.255.255 port 67
Dec 12 19:57:19 DietPi dhclient[459]: DHCPREQUEST for 192.168.31.238 on eth0 to 255.255.255.255 port 67
Dec 12 19:57:19 DietPi ifup[459]: DHCPREQUEST for 192.168.31.238 on eth0 to 255.255.255.255 port 67
Dec 12 19:57:19 DietPi dhclient[459]: DHCPACK of 192.168.31.238 from 192.168.31.1
Dec 12 19:57:19 DietPi ifup[459]: DHCPACK of 192.168.31.238 from 192.168.31.1
Dec 12 19:57:20 DietPi dhclient[459]: bound to 192.168.31.238 -- renewal in 18514 seconds.
Dec 12 19:57:20 DietPi ifup[459]: bound to 192.168.31.238 -- renewal in 18514 seconds.
Dec 12 19:57:20 DietPi systemd[1]: Finished ifup for eth0.
HA Supervised not installed.
As stated above, NetworkManager is conflicting with ifupdown. You would need to choose which one to use and remove the other one. Keep in mind, if you remove ifupdown, you will lose possibility to manage network connections via dietpi-config.
Anyway, this might not solve the 10 second delay during boot, but would be good to clean it up.
One more question stepped into my mind. Is there a specific service that is actually delayed? Because I don’t see anything. Other services continue to start in parallel, event that ifup@eth0.service did not finished.
