This website is deprecated, and remains online only for historic access to old issues and docs for historic versions of Virtualmin. It has been unmaintained for several years, and should not be relied on for up-to-date information. Please visit www.virtualmin.com instead.
Getting a new VPS for work next week. I ran VirtualMin on CentOS7 for my personal use, but would love to start the new VPS with Ubuntu 18.04.
Have things been fixed from VirtualMin side? I know it’s a bad idea to run company internal stuff on Ubuntu anyways, but would like to use something different than CentOS :)
So i am on Digital Ocean (DO) with no backup, no ipv6, no script, yes monitor.
The person on comment #31 made me trying to manually setting netplan. I read /etc/netplan/50-cloud-init.yaml but my list of ip addresses is not quite like the person in the comment have mentioned so I simply try my luck and just run sudo netplan apply and it throw me error about set-name is wrong or something like that. And this is my yaml file.
Referring to the man page here http://manpages.ubuntu.com/manpages/cosmic/man5/netplan.5.html#description about set-name seems that it has the wrong indentation, so I moved the set-name line 1 level up, which is the same level as match. Redo sudo netplan apply and this time it gives no error and be able to access the host again via ssh. Shutdown, reboot etc is also fine, too.
Of course I can only speak for myself and not sure if this is a valid fix and/or can apply to other people. It will be good if something from virtualmin can give this a check.
It's expected to work now that Webmin 1.900 is in the repos, but it's untested by me, so I haven't added it back to the supported OS list yet. I welcome success/fail reports from folks who want to try it (file a ticket or start a new thread in the forum about it, don't keep this old thread alive forever).
Hi there,
I just want to confirm that Virtualmin are now support for Ubuntu 18.04 LTS in install.sh even the supported OS haven't listed yet.
I'm using Webmin 1.900. No more network problem after rebooting server. I'm able to ssh to server as usual. I use DigitalOcean.
Thank you.
Welcome to Ubuntu 16.04.5 LTS (GNU/Linux 4.17.17-x86_64-linode116 x86_64)
New release '18.04.1 LTS' available.
Run 'do-release-upgrade' to upgrade to it.
and Virttualmin is saying:
Recent package updates (such as a new kernel version) require a reboot to be fully applied
Are these related and do you think I am reasonably safe updating a production server to 18.04.1 LTS at this time?
The Virtualmin message is unrelated to the server message.
The latter message from Virtualmin is saying something has changed that wants a reboot (like a kernel upgrade). That's something that already happened; you have already upgraded whatever package it is that wants a reboot. You just need to reboot to get the new kernel, or whatever. That's a minor thing, assuming you didn't enable any weird third party repos or anything. Just reboot. It'll be fine.
The server message is saying there is a distribution upgrade available to 18.04.1. That would be a big deal, regardless of Virtualmin being involved. Upgrades are expected to work, because upgrading does not change the network configuration system (so it'll be the old one, which has no known issues, or at least, no serious ones). But, upgrading the distribution is something that should not be embarked upon lightly. It will break your web applications, and will need to tweaking to bring them back. Virtualmin can usually fix some of it...just do a "re-check configuration", so it can learn about the new PHP version, and update all the config files, but don't plan on it...expect the worst, and you'll be pleased when it goes easier than expected, rather than expecting an OS major version change to be completely free of problems. But, databases might need attention, etc.
And, if you do a system upgrade, you must have good backups of everything.
In short: Updates are minor, do them every day. An OS upgrade is major, do it only after preparing for the worst and only if it will give you something you need (like a new version of PHP or something that you need for a specific application).
I've gotten Ubuntu 18.04.1 LTS to work on my DigitalOcean droplet, but with teething issues:
- Nginx doesn't run automatically upon reboot; needed to change some config settings to work
- DKIM doesn't currently work https://www.virtualmin.com/node/59560
The latest Virtualmin works fine with Ubuntu 18.04.1 LTS so far, however install.sh script still don't state about supporting 18.04.1 LTS.
UPDATE, day after.
DO NOT INSTALL VIRTUALMIN ON PRODUCTION SERVER YET, AS IT ISN'T SUPPORTED. I spent all night trying to fix the Network Interfaces. If you have only 1 IP that's probably fine, but I have a lot IPs and there was a problem with Network Configuration...
First put your netplan config in "code" brackets. The spacing and hierarchy matters. What you pasted is useless for diagnosing.
Second, when running netplan apply, it tells you the line number and column where the errors are. Open the config file, space down and over that number and see whats wrong.
But all you need is IP and gateway. This should get you going. Although I think that /24 on the ip should be /32, that is unless you where allotted all 254 addresses of the 207.180.193.x block. It can always be /32. Note spacing below must be spaces not tabs and hierarchy matters. Also note, quotes around IPV6 addresses are required.
network: version: 2 ethernets: eth0: dhcp4: no dhcp6: no addresses: [207.180.193.199/32] gateway4: 207.180.193.1 nameservers: addresses: [127.0.0.1,"2606:4700:4700::1111","2606:4700:4700::1001",1.1.1.1,1.0.0.1]
I dont know what those "invalid addresses" you listed are for as they are not even in your IP subnets. But get ipv4 going, then add ip6 stuff. Should only require additional ip and gateway. Of which gateway6: fe80::1 is wrong. Fe80 = localhost and a gateway cant be localhost.
However, that was not the whole fix sadly. When I do this, after and then reboot VPS, I can't connect either via browser, ssh or ftp and 'virtualmin check-config' returns :
'Virtualmin could not work out the default IPv4... '
Hi, no I used it as a boilerplate but kept all other data from my own file. On my old file I had eth0 so I kept it. Anyway, I will double check and try once more and if I have problems I will probably quit this and install Centos 7 instead. :) Thanks for the reply!
The first time I rebooted, the network was down (I had to access through VNC from my VPS provider), but tweaking above settings (I had a /36 prefix for IPv4 address, instead of /24 for example, or forgot to put the right interface name, ...) copying them from the previous default .yaml file, but keeping the exact structure worked really fine, and I'm right into Virtualmin now.
Actually, it seems Virtualmin continues to aggressively change the netplan conf.
After the installation, I find the nameservers wrongly set:
# netplan apply /etc/netplan/01-netcfg.yaml:9:28: Error in network definition: malformed address ' 213.136.95.11', must be X.X.X.X or X:X:X:X:X:X:X:X addresses: [127.0.0.1,213.136.95.10,' 213.136.95.11']
I had to go back and delete the single quotes around that IP and the preceeding space.
Using the script on Ubuntu 18.04 I'm not able to manage the emails to sign with DKIM, any of you think is a bug for using the latest LTS of ubuntu and the script?
Does that work for you?
Thanks.
Update: LOG: warning: connect to Milter service inet:localhost:8891: Connection refused
Because I'm a bit of a nut when it comes to dns stuff, come to find out, postfix was sending out the wrong banner. It didn't update to the new hostname, so I went into
Webmin->Servers->Postfix Mail Server->General Options and set
Internet hostname of this mail system to Default (provided by system)
Local internet domain name Default (provided by system)
Restarted Postfix and the SMTP banner is now correct and looking back in General Options, it's back to a custom setting with the right info this time.
I confirm that Virtualmin messes up /etc/netplan/01-netcfg.yaml on a fresh Ubuntu 18.04.2 LTS.
Perhaps it only happens when having multiple IPs, as I have both IPv4 and IPv6 on mine.
Getting a new VPS for work next week. I ran VirtualMin on CentOS7 for my personal use, but would love to start the new VPS with Ubuntu 18.04.
Have things been fixed from VirtualMin side? I know it’s a bad idea to run company internal stuff on Ubuntu anyways, but would like to use something different than CentOS :)
Or go with good old CentOS?
Thanks guys :)
I just set up another production server last week -- used DEBIAN 9...
Smooth as silk.
Thanks jimdunn, but this thread is all about Ubuntu 18.04
I've been installing Virtualmin on Centos for over 10 years now. I did two last week. But, I wanted to setup an 18.04 to test TYPO3 9.4
So i am on Digital Ocean (DO) with no backup, no ipv6, no script, yes monitor.
The person on comment #31 made me trying to manually setting netplan. I read
/etc/netplan/50-cloud-init.yaml
but my list of ip addresses is not quite like the person in the comment have mentioned so I simply try my luck and just runsudo netplan apply
and it throw me error aboutset-name
is wrong or something like that. And this is my yaml file.network:
version: 2
ethernets:
eth0:
addresses: [MULTIPLE_IP_ADDRESSES]
gateway4: IP_ADDRESS
nameservers:
addresses: [MULTIPLE_IP_ADDRESSES]
match:
macaddress: MAC_ADDRESS
set-name: eth0
Referring to the man page here http://manpages.ubuntu.com/manpages/cosmic/man5/netplan.5.html#description about
set-name
seems that it has the wrong indentation, so I moved theset-name
line 1 level up, which is the same level asmatch
. Redosudo netplan apply
and this time it gives no error and be able to access the host again via ssh. Shutdown, reboot etc is also fine, too.Of course I can only speak for myself and not sure if this is a valid fix and/or can apply to other people. It will be good if something from virtualmin can give this a check.
Hello, I am wondering what the current status is with virtualmin/webmin and Ubuntu 18? Does it work now or should we stick with 16LTS for now?
I didn't see Ubuntu 18 listed on the virtualmin page but I just went through an install and it rebooted. Everything seems fine.
Thanks
It's expected to work now that Webmin 1.900 is in the repos, but it's untested by me, so I haven't added it back to the supported OS list yet. I welcome success/fail reports from folks who want to try it (file a ticket or start a new thread in the forum about it, don't keep this old thread alive forever).
--
Check out the forum guidelines!
Hi there, I just want to confirm that Virtualmin are now support for Ubuntu 18.04 LTS in install.sh even the supported OS haven't listed yet. I'm using Webmin 1.900. No more network problem after rebooting server. I'm able to ssh to server as usual. I use DigitalOcean. Thank you.
hmm, ok, my server is saying:
and Virttualmin is saying:
Are these related and do you think I am reasonably safe updating a production server to 18.04.1 LTS at this time?
Chris
The Virtualmin message is unrelated to the server message.
The latter message from Virtualmin is saying something has changed that wants a reboot (like a kernel upgrade). That's something that already happened; you have already upgraded whatever package it is that wants a reboot. You just need to reboot to get the new kernel, or whatever. That's a minor thing, assuming you didn't enable any weird third party repos or anything. Just reboot. It'll be fine.
The server message is saying there is a distribution upgrade available to 18.04.1. That would be a big deal, regardless of Virtualmin being involved. Upgrades are expected to work, because upgrading does not change the network configuration system (so it'll be the old one, which has no known issues, or at least, no serious ones). But, upgrading the distribution is something that should not be embarked upon lightly. It will break your web applications, and will need to tweaking to bring them back. Virtualmin can usually fix some of it...just do a "re-check configuration", so it can learn about the new PHP version, and update all the config files, but don't plan on it...expect the worst, and you'll be pleased when it goes easier than expected, rather than expecting an OS major version change to be completely free of problems. But, databases might need attention, etc.
And, if you do a system upgrade, you must have good backups of everything.
In short: Updates are minor, do them every day. An OS upgrade is major, do it only after preparing for the worst and only if it will give you something you need (like a new version of PHP or something that you need for a specific application).
--
Check out the forum guidelines!
Joe,
Excellent info, thank you!
Chris
I've gotten Ubuntu 18.04.1 LTS to work on my DigitalOcean droplet, but with teething issues: - Nginx doesn't run automatically upon reboot; needed to change some config settings to work - DKIM doesn't currently work https://www.virtualmin.com/node/59560
The latest Virtualmin works fine with Ubuntu 18.04.1 LTS so far, however install.sh script still don't state about supporting 18.04.1 LTS.
UPDATE, day after.
DO NOT INSTALL VIRTUALMIN ON PRODUCTION SERVER YET, AS IT ISN'T SUPPORTED. I spent all night trying to fix the Network Interfaces. If you have only 1 IP that's probably fine, but I have a lot IPs and there was a problem with Network Configuration...
Where can I find the beta script for installing on Ubuntu 18.04?
Eugene van der Merwe
https://vander.host WordPress Website Hosting, VPS Hosting, and Domain Registration
Hi,
I installed a fresh 18.04.2 on a VPS, and set netplan as follows:
network:
version: 2
renderer: networkd
ethernets:
ens18:
match:
macaddress: "00:50:56:3c:f8:78"
set-name: ens18
addresses:
- 207.180.193.199/24
- 2a02:c207:2022:2714:0000:0000:0000:0001/64
gateway4: 207.180.193.1
gateway6: fe80::1
nameservers:
search: [ invalid ]
addresses:
- 213.136.95.10
- 213.136.95.11
- 2a02:c207::1:53
- 2a02:c207::2:53
Then I try installing virtualmin GPL, and it throws a post-installation error on network. Looking now at netplan's config:
network:
version: 2
renderer: networkd
ethernets:
ens18:
addresses: ['/32']
nameservers:
addresses: [127.0.0.1,127.0.0.53]
search: [invalid]
match:
macaddress: "00:50:56:3c:f8:78"
macaddress: "00:50:56:3c:f8:78"
set-name: ens18
addresses:
- 207.180.193.199/24
- 2a02:c207:2022:2714:0000:0000:0000:0001/64
gateway4: 207.180.193.1
gateway6: fe80::1
nameservers:
search: [ invalid ]
addresses:
- 213.136.95.10
- 213.136.95.11
- 2a02:c207::1:53
- 2a02:c207::2:53
It has new lines, not needed, and throwing errors with #netplan apply
Is there a way to make it properly work?
Thank you,
Nicola Urbinati Dreamlord Press www.dreamlordpress.it
First put your netplan config in "code" brackets. The spacing and hierarchy matters. What you pasted is useless for diagnosing. Second, when running netplan apply, it tells you the line number and column where the errors are. Open the config file, space down and over that number and see whats wrong.
But all you need is IP and gateway. This should get you going. Although I think that /24 on the ip should be /32, that is unless you where allotted all 254 addresses of the 207.180.193.x block. It can always be /32. Note spacing below must be spaces not tabs and hierarchy matters. Also note, quotes around IPV6 addresses are required.
network:
version: 2
ethernets:
eth0:
dhcp4: no
dhcp6: no
addresses: [207.180.193.199/32]
gateway4: 207.180.193.1
nameservers:
addresses: [127.0.0.1,"2606:4700:4700::1111","2606:4700:4700::1001",1.1.1.1,1.0.0.1]
I dont know what those "invalid addresses" you listed are for as they are not even in your IP subnets. But get ipv4 going, then add ip6 stuff. Should only require additional ip and gateway. Of which gateway6: fe80::1 is wrong. Fe80 = localhost and a gateway cant be localhost.
Hi,
Thank you, I edited my previous comment with code brackets to make it readable.
My problem is:
I'll try your much simpler config structure.
Nicola Urbinati Dreamlord Press www.dreamlordpress.it
Well, with this netplan config, everything works ok.
network:
version: 2
renderer: networkd
ethernets:
ens18:
addresses: [207.180.193.199/24, '2a02:c207:2022:2714::1/64']
gateway4: 207.180.193.1
gateway6: 'fe80::1'
nameservers:
search: [ invalid ]
addresses: [213.136.95.10, 213.136.95.11, '2a02:c207::1:53', '2a02:c207::2:53']
Thank you very much.
Nicola Urbinati Dreamlord Press www.dreamlordpress.it
Thanks, I was trying to do an automated install on fresh Ubuntu 18.04 but I was getting error Net configuration: No interface named address found.
Modifying netplan config to this solved the error.
However, that was not the whole fix sadly. When I do this, after and then reboot VPS, I can't connect either via browser, ssh or ftp and 'virtualmin check-config' returns : 'Virtualmin could not work out the default IPv4... '
Did you copy/paste the script as-is? Your interface (ens18 here) could be different, or some other detail specific to your installation.
Nicola Urbinati Dreamlord Press www.dreamlordpress.it
Hi, no I used it as a boilerplate but kept all other data from my own file. On my old file I had eth0 so I kept it. Anyway, I will double check and try once more and if I have problems I will probably quit this and install Centos 7 instead. :) Thanks for the reply!
I'm actually doing it once more as well in a few minutes (as soon as the VPS finishes to reinstall), so let's see if it's stable.
Nicola Urbinati Dreamlord Press www.dreamlordpress.it
Ok,
You have to put the right prefix (my new machine wants /24), and all things needed by your particular setup, but the structure is solid:
[SUCCESS] Installation Complete!
[SUCCESS] If there were no errors above, Virtualmin should be ready
Nicola Urbinati Dreamlord Press www.dreamlordpress.it
Thanks for taking the time to reply.
Did you also try to restart the server after you had it installed? In my case, it worked but when I restarted, it never came back online...
Hi,
Yes, I did.
The first time I rebooted, the network was down (I had to access through VNC from my VPS provider), but tweaking above settings (I had a /36 prefix for IPv4 address, instead of /24 for example, or forgot to put the right interface name, ...) copying them from the previous default .yaml file, but keeping the exact structure worked really fine, and I'm right into Virtualmin now.
Nicola Urbinati Dreamlord Press www.dreamlordpress.it
Hi,
Actually, it seems Virtualmin continues to aggressively change the netplan conf.
After the installation, I find the nameservers wrongly set:
# netplan apply
/etc/netplan/01-netcfg.yaml:9:28: Error in network definition: malformed address ' 213.136.95.11', must be X.X.X.X or X:X:X:X:X:X:X:X
addresses: [127.0.0.1,213.136.95.10,' 213.136.95.11']
I had to go back and delete the single quotes around that IP and the preceeding space.
Hope doesn't get messed up again in the future.
Nicola Urbinati Dreamlord Press www.dreamlordpress.it
Hi there,
Using the script on Ubuntu 18.04 I'm not able to manage the emails to sign with DKIM, any of you think is a bug for using the latest LTS of ubuntu and the script?
Does that work for you?
Thanks.
Update: LOG: warning: connect to Milter service inet:localhost:8891: Connection refused
Update 2: With help of @brad100 I solved the problem https://www.virtualmin.com/comment/808804#comment-808804 Also found helpful: https://www.virtualmin.com/node/59560
Had the same issue, thanks @carras and @brad100 for solving this problem.
Because I'm a bit of a nut when it comes to dns stuff, come to find out, postfix was sending out the wrong banner. It didn't update to the new hostname, so I went into
Webmin->Servers->Postfix Mail Server->General Options and set
Internet hostname of this mail system to Default (provided by system)
Local internet domain name Default (provided by system)
Restarted Postfix and the SMTP banner is now correct and looking back in General Options, it's back to a custom setting with the right info this time.
I confirm that Virtualmin messes up /etc/netplan/01-netcfg.yaml on a fresh Ubuntu 18.04.2 LTS.
Perhaps it only happens when having multiple IPs, as I have both IPv4 and IPv6 on mine.
Here's my original, working 01-netcfg.yaml:
network:
version: 2
renderer: networkd
ethernets:
enp2s0:
addresses:
- MY_IPv4/32
- MY_IPv6/64
routes:
- on-link: true
to: 0.0.0.0/0
via: MY_IPv4_GATEWAY
gateway6: fe80::1
nameservers:
addresses:
- 1.1.1.1
- 8.8.8.8
Here it is right after installing Virtualmin 6.07, which makes network connection fail on boot (I think IPv6 still worked):
network:
version: 2
renderer: networkd
ethernets:
enp2s0:
addresses: ['MY_IPv4/32','MY_IPv6/64']
gateway6: fe80::1
nameservers:
addresses: [127.0.0.1,1.1.1.1,8.8.8.8]
routes:
to: 0.0.0.0/0
via: MY_IPv4_GATEWAY
(Placeholders: MY_IPv4, MY_IPv6, MY_IPv4_GATEWAY)
Pages