Proxmox VE 6.0 beta released!

Shusky

Member
Feb 19, 2018
11
3
8
24
Munich, Germany
Just successfully performed the upgrade from the latest 5.4 version to 6.0.

The only thing that came to my attention during the installation was that grub did not recognize the grub installed on my two S-ATA SSDs, which are running in a zpool mirror and contain the Proxmox installation. After manually selecting those drives once again, the upgrade continued normally and there were no issues during the reboot.

Hardware:
CPU: Intel Xeon E5-2630 v4 (40) @ 2.2 GHz
MB: Supermicro X10DRL-i - BIOS 3.1c
RAM: 8x Samsung M393A4K40BB0-CPB
SSD: 2x Samsung 860 EVO 250GB, 4x Samsung 970 EVO 500GB, 1x Intel Optane 900P 280GB
 
  • Like
Reactions: fireon

ivensiya

Active Member
Feb 17, 2013
60
1
28
Openvswitch not working. When node is up, i can see only local. Same config i used at another node with 5.4 pve, no probem.


auto lo
iface lo inet loopback

allow-vmbr0 eth0
iface eth1 inet manual
ovs_type OVSPort
ovs_bridge vmbr0

auto eth0.4003
iface eth0.4003 inet static
address 192.168.200.115
netmask 255.255.255.0
vlan-raw-device eth0
mtu 1400

auto vmbr0

iface vmbr0 inet static
address X.X.X.X
netmask 255.255.255.192
gateway X.X.X.X
ovs_type OVSBridge
ovs_ports eth0
 

t.lamprecht

Proxmox Staff Member
Staff member
Jul 28, 2015
5,214
1,520
164
South Tyrol/Italy
shop.proxmox.com
It simply took me to a screen, that said “Reboot into firmware interface”.

However, when I re-did the installation using ext4 - I was able to boot successfully
Will a UEFI installation be possible in the future?
Please use the debug mode of the installer, at the second prompt check for /sys/firmware/efi directory, if it is not present or empty you are not booted in uefi mode. I'll add a line at the summary screen to show which boot mode the installer is in, it seems that it may help some people to ensure that they have correct the mode selected.

If you're booted into uefi mode and choose ZFS everything should be done automatically.
 
Last edited:

dcsapak

Proxmox Staff Member
Staff member
Feb 1, 2016
7,874
952
163
33
Vienna
Openvswitch not working. When node is up, i can see only local. Same config i used at another node with 5.4 pve, no probem.
maybe the name of the network name has changed? check the dmesg output
 

casparsmit

Active Member
Feb 24, 2015
38
1
28
This seems rather odd (segfaults in core binaries usually makes me think about broken RAM).
Please open a separate thread, since this seems a more involved issue, and provide the following:
* the diff of the shipped udev init-script and the one you have on your system (did you modify it)?
* what is written in the systemd-override-conf for udev ('/etc/systemd/system/systemd-udevd.service.d/override.conf')?
* the output of `dmesg`
* the journal since the last boot `journalctl -b`

Thanks!

Thanks for the hints! It turned out i had /etc/systemd/system/systemd-udevd.service.d/override.conf configured only containing:

[Service]
ExecStartPost=/bin/sleep 15

Removing this file before the upgrade fixed the problem. I didn't expect this small change could potentially lead to a broken system.
Thanks again, now to test PVE6 :)
 

sienar

Active Member
Jul 19, 2017
48
8
28
42
Awesome, super psyched to get this newer version of ZFS upgrade. But before I try on my lone single server, anyone have success upgrading a machine with ZFS on root?
 

Breymja

Member
Aug 14, 2017
58
1
13
31
Please use the debug mode of the installer, at the second prompt check for /sys/firmware/efi directory, if it is not present or empty you are not booted in uefi mode. I'll add a line at the summary screen to show which boot mode the installer is in, it seems that it may help some people to ensure that they have correct the mode selected.

If you're booted into uefi mode and choose ZFS everything should be done automatically.

That unfortunately does not work.
I boot via the UEFI option on the installation media, the folder shows and has contents. After DHCP it tells me that the installations is aborted and cannot continue. That's it, no more information given. If i choose the legacy boot option i can install Proxmox just fine, but don't have a UEFI installation. That's what i meant. So ZFS booting from a RAID1 of NVME SSDs = Works. Installing the same in UEFI mode = Does not work.

See attached screenshots.
proxmox.png
proxmox2.png
proxmox3.png
 

t.lamprecht

Proxmox Staff Member
Staff member
Jul 28, 2015
5,214
1,520
164
South Tyrol/Italy
shop.proxmox.com
After DHCP it tells me that the installations is aborted and cannot continue

Strange! Could you switch over to the second tty, normally done by sending "CTRL + ALT + F2" (use F2 to switch back to the original prompt) and check if there are any errors visible about why this happens. Note that sending this combination is best done via the virtual keyboard of the remote viewer, at least if you use Linux as Desktop - as else your local Desktop will switch to TTY2, (and the xserver is normal on TTY7).
 

fabian

Proxmox Staff Member
Staff member
Jan 7, 2016
7,483
1,396
164
@Breymja there is a udevd worker getting killed in your screenshot - maybe you could provide censored logs from inside the installer? e.g., dmesg? you should be able to use the shell to dump it somewhere (external media, network share, ...)
 

Shusky

Member
Feb 19, 2018
11
3
8
24
Munich, Germany
Awesome, super psyched to get this newer version of ZFS upgrade. But before I try on my lone single server, anyone have success upgrading a machine with ZFS on root?
Also need to do this on 2 machines, and would prefer to read some success stories first...:oops:
As mentioned in Post #23, I successfully updated my Proxmox machine with ZFS on root.
 

Attachments

  • PVE6_rpool.png
    PVE6_rpool.png
    53.1 KB · Views: 20

Breymja

Member
Aug 14, 2017
58
1
13
31
Strange! Could you switch over to the second tty, normally done by sending "CTRL + ALT + F2" (use F2 to switch back to the original prompt) and check if there are any errors visible about why this happens. Note that sending this combination is best done via the virtual keyboard of the remote viewer, at least if you use Linux as Desktop - as else your local Desktop will switch to TTY2, (and the xserver is normal on TTY7).
I'd love to do that, but the shortcut does not work with the LARA from Hetzner. I can neither do it via my keyboard nor via the softkeyboard. Nothing happens.

@Breymja there is a udevd worker getting killed in your screenshot - maybe you could provide censored logs from inside the installer? e.g., dmesg? you should be able to use the shell to dump it somewhere (external media, network share, ...)
I attached the dmesg.log, but it's really hard to get that with the slowlyness and english keyboard.
 

Attachments

  • dmesg.log
    52.8 KB · Views: 7

Breymja

Member
Aug 14, 2017
58
1
13
31
Hetzner support told me that Debian 9 Kernel does not support the NIC in EX62 Servers. For Debian 10 it should work and i have internet, as via that i did transfer the dmesg.log file. Other than that, they also have no idea what the issue could be.
 

Breymja

Member
Aug 14, 2017
58
1
13
31
For some reason, after Hetzner inserted a USB Stick directly the installation also came up in UEFI mode. (Seems there are timeouts when using virtual media on uefi)

I can now either boot via uefi and efibootmgr works and /sys/firmware/efi is there or i can boot without, then it boots via grub. Is that intended?
 
Last edited:
May 18, 2019
167
8
23
Los Angeles, CA USA
Q: Can I install Proxmox VE 6.0 beta on top of Debian Buster?
A: Yes, see https://pve.proxmox.com/wiki/Install_Proxmox_VE_on_Debian_Buster

Q: Can I dist-upgrade Proxmox VE 5.4 to 6.0 beta with apt?
A: Please follow the upgrade instructions exactly, as there is a major version bump of corosync (2.x to 3.x)
https://pve.proxmox.com/wiki/Upgrade_from_5.x_to_6.0

I have a single node Debian 9.9 on UEFI boot. No Ceph, no Corosync, boot is on ext4 (mdadm Raid 1) spinning disk. PVE is current.

Upgraded following instructions. startup seems fine. web gui seems fine. guests (KVM & LXC) seem fine, the console comes up.

but no proxmox cmds work (pveeam, pct, qm) work. uname -a shows 5.0.15-1-pve #1 SMP PVE 5.0.15-1.

The binaries are there, /usr/sbin/pct enter VMID works.

Code:
echo $PATH
/usr/local/bin:/usr/bin:/bin:/usr/local/games:/usr/games

is this a sign that there is something else wrong or I just need to adjust PATH, and if so, whats the proper way to update PATH?
 

Stoiko Ivanov

Proxmox Staff Member
Staff member
May 2, 2018
6,862
1,039
164
  • Like
Reactions: Gaia

About

The Proxmox community has been around for many years and offers help and support for Proxmox VE, Proxmox Backup Server, and Proxmox Mail Gateway.
We think our community is one of the best thanks to people like you!

Get your subscription!

The Proxmox team works very hard to make sure you are running the best software and getting stable updates and security enhancements, as well as quick enterprise support. Tens of thousands of happy customers have a Proxmox subscription. Get your own in 60 seconds.

Buy now!