Yeah, that is exactly what we were doing. We used ZFS to do replication of the docker-in-LXC services between servers. I guess we need to rebuild that in qemu now, but docker-in-lxc on ZFS, even with AUFS was just super easy in terms of configuration, maintenance, replication. It even enabled a pretty useful sort-of-HA, because */15 replication only takes 3 seconds and LXC startup time is 5-10 seconds.
Proxmox VE 7.0 released!
- Thread starter martin
- Start date
Correct. Blue screen boot selection if fine, text during boot up is just fine, but the video mode change before the login screen switches to a bad video mode.
Can ssh with no problems, can get to the web console with no problems, can join a cluster with no problems.
Thanks!
I just wonder if Proxmox staff will somewhen realize native Web-interface controlled Docker support in PVE (in addition to LXC and QEMU)?
Last edited:
Ok, thanks for confirmation.
Great that at least this all is working, so it's really only the graphic driver that breaks.
Can you please open a new thread, tag my username with @t.lamprecht and provide some more info, for now the output of
lspci -knn and it would be great if you can check the journal (or dmesg) for any error regarding display/drm/gpu or the like and post that too.
No plans there, I'm afraid. Also, this is the thread for the Proxmox VE 7.0 release
FYI: there's some progress going on in the overlayfs support on ZFS, which would be great in general, but one cannot really guess yet when it'll be actually shipped. PVE 6.4 will be still supported for almost a year, so you could wait at least a few months to see how that plays out before investing in change to VMs there, in the worst case you still have the same amount of work as you'd have now.
FWIW, you could also build the module yourself: https://github.com/sfjro/aufs5-standalone/tree/aufs5.11 but that is definitively more involved.
Upgraded a 4 node production cluster running Ceph with no downtime or issues. I did do the needful and ran the check script, ensured Prox 6.4 was fully updated, and reviewed my systems for any of the known issues/extra steps. For example, 3/4 nodes in this cluster run old boards and therefore can only BIOS boot, not EFI boot, so I made sure their ZFS boot pools were properly switched to the proxmox boot method. I have not yet upgraded Ceph to Pacific due to time constraints.
10/10 major version release. Lots of nice little UI improvements, some apparent performance improvements on my hardware, and continued support for hardware that should have been recycled years ago.
10/10 major version release. Lots of nice little UI improvements, some apparent performance improvements on my hardware, and continued support for hardware that should have been recycled years ago.
This is not a error but a warning that you're about to upgrade to a new major release.
It's there to ensure people are aware of the upgrade how-to and that a major upgrade is about to start.
The
pve-manager (currently 6.4-13) package and the proxmox-ve (currently indeed version 6.4-1) meta package are two different ones, here the latter is used.If you're aware of the upgrade how-to and ready to upgrade you can press enter to continue
Hi,
Congratulations on the new release!
I have a fresh PVE 7 installation for testing and manually adding VLAN interface (vmbr0.1) on the bridge (vmbr0) like I did on PVE 6.3 does not work. Adding "hwaddress" option to them helped (I have not tested if it's needed for both or maybe just one of them), even if our network is not MAC-restricted. Hint: reboot was needed in my case (after testing different network settings).
It looks to me that:
- showing MAC in the GUI makes sense
- adding hwaddress by the installer would be nice
- it would be great if PVE could offer VLAN settings at the installation time to avoid error-prone manual setup.
Congratulations on the new release!
I have a fresh PVE 7 installation for testing and manually adding VLAN interface (vmbr0.1) on the bridge (vmbr0) like I did on PVE 6.3 does not work. Adding "hwaddress" option to them helped (I have not tested if it's needed for both or maybe just one of them), even if our network is not MAC-restricted. Hint: reboot was needed in my case (after testing different network settings).
It looks to me that:
- showing MAC in the GUI makes sense
- adding hwaddress by the installer would be nice
- it would be great if PVE could offer VLAN settings at the installation time to avoid error-prone manual setup.
Last edited:
This was already reported several times - you just need to clear the browser cache and reload the page.
Yes, and there are many place with issues. These are the reasons why we do not recommend btrfs for now, its just a preview technology in Proxmox VE.
Nice to see the correct memory usage in the Dashboard ! Thanks !
(LXC = correct values, VM = still not correct values).
(LXC = correct values, VM = still not correct values).
For VMs ensure that the VirtIO balloon driver is installed so that PVE actually gets the correct values reported from the OS running in the VM. For example, if Windows is running as OS then check https://pve.proxmox.com/wiki/Windows_VirtIO_Drivers