How can I create a start up script for proxmox for when the system boots... e.g.: I would like to update "ethtool -K eno1 tso off gso off" to be written each time the system starts
Start-up/boot script
- Thread starter friendodevil
- Start date
Option 1.) You can add a cron line (
Option 2.) You create a oneshot systemd service like described here: https://trstringer.com/simple-vs-oneshot-systemd-service/
Option 3.) In case it is network related you also might want to use the "pre-up", "post-up" and so on constructs in your /etc/network/interfaces file. Adding a line like
So for your specific example I would prefer option 3. For simple stuff that just has to run once but its not important when, I would use option 1. Option 2 is the most powerful but also the most complex one that for example would allow you at which point in time in the boot sequence you want your command to be run.
crontab -e) like this: @reboot /usr/sbin/ethtool -K eno1 tso off gso off > /dev/null 2>&1.Option 2.) You create a oneshot systemd service like described here: https://trstringer.com/simple-vs-oneshot-systemd-service/
Option 3.) In case it is network related you also might want to use the "pre-up", "post-up" and so on constructs in your /etc/network/interfaces file. Adding a line like
pre-up /usr/sbin/ethtool -K eno1 tso off gso off below your eno1 interface ( iface eno1 inet ...) should run that command each time before bringing the eno1 NIC up.So for your specific example I would prefer option 3. For simple stuff that just has to run once but its not important when, I would use option 1. Option 2 is the most powerful but also the most complex one that for example would allow you at which point in time in the boot sequence you want your command to be run.
Last edited:
I did option 3 to run "pre-up pvecm expected 1" but for some reason it does not work. Here is my file :
auto lo
iface lo inet loopback
iface eno1 inet manual
pre-up pvecm expected 1
iface eno3 inet manual
iface eno4 inet manual
iface enp5s0 inet manual
iface eno2 inet manual
auto vmbr0
iface vmbr0 inet static
address 192.168.0.102/24
gateway 192.168.0.1
bridge-ports eno1
bridge-stp off
bridge-fd 0
What am I doing wrong?
First question would be why you want to set "pvecm expected 1" at all? This is dangerous and shouldn't be run as default. You should only manually run this after verifying that the other node is dead or you might run into a dangerous split brain situation. Proper solution when running a 2 node cluster would be to get something to act as a qdevice for the third vote so you don't have to temper with "pvecm expected 1" when shutting down one node of your two node cluster: https://pve.proxmox.com/wiki/Cluster_Manager#_corosync_external_vote_support
During the winter here I only run 1 VM that is not that critical and I have 3 node setup. For like 6 months it will be like that and I want to save power since my nodes are actual 4U servers and consume a lot of power and this place only has solar power during the winter so I need to run it like that during winter.
Please help me run this on every startup.
Then I would try option 1:
So:
@reboot /usr/bin/pvecm expected 1 > /dev/null 2>&1 to your crontab.
Last edited:
I run this command and just now power went off and on again and seems it did not work since I had to manually run the command again to run the VM...
Ok so Here's a step by step guide:
1. You open the shell in the required node and run the command (crontab -e)
2. You will probably be asked to choose a text editor, choose nano
3. Scroll all the way down in the opened file and add this line: @reboot /usr/bin/pvecm expected 1 > /dev/null 2>&1
4. Save the File
5. Reboot
I came across this thread from google. My use case is that I only use enabled clustering for a single management UI. I don't use any clustering features.
I tried the suggestion of adding a reboot command to crontab. But, it looks like the reboot command runs before proxmox is able to accept the pvecm command.
The solution that worked for me was adding a sleep prior to the pvecm command as below. 300s worked for me.
I tried the suggestion of adding a reboot command to crontab. But, it looks like the reboot command runs before proxmox is able to accept the pvecm command.
The solution that worked for me was adding a sleep prior to the pvecm command as below. 300s worked for me.
@reboot /bin/sleep 300 ; /usr/bin/pvecm expected 1 > /dev/null 2>&1