Search results

All our containers include /usr/share/lxc/config/common.conf which in turn includes all .conf files in /usr/share/lxc/config/common.conf.d/ - You can add eg. a file 99-pid-limit.conf to that directory and add the above cgroup line to it.

Also note the related entry here[1]. The default size is a guess made by the LVM tools, and was updated by us on our ISO after users started having issues with that. [1] https://bugzilla.proxmox.com/show_bug.cgi?id=1241

But there the disk had `backup=0` when the error was posted.

Well, pve-zsync is supposed to ignore the the `format=` option in the config. Could you post the error messages you see along with the disks found in the VM's config at the time of a failure?

They ended up with a kernel-side fix for this. There's a kernel in the `pvetest` repository ready to be tested. Package pve-kernel-4.15.18-1-pve version 4.15.18-17.

Auf pvetest is ein Kernel-Paket (Version 4.15.18-17) mit dem obigen fix. Wäre nett, wenn du ihn testen könntest.

Geht natürlich auch ohne systemd, je nachdem was drin tatsächlich passiert. Die Meldung sieht schon stark nach demselben Problem aus. Zumindest gibt's nur 1 Operation die das auslösen kann. Sollte also schon auch zusammen mit diesem bug[1] gefixt werden, hoffe ich mal. [1]...

The corosync service uses /etc/corosync/corosync.conf, you need to make sure the same changes are applied there. This file is usually updated by the cluster service when the corosync.conf in /etc/pve gets updated by one of the nodes, but if you make manual changes, you need to make sure they're...

Looks like the pve-cluster service is not running properly. Assert this by stopping it (`systemctl stop pve-cluster.service`). Then run `pmxcfs -l` to start the cluster file system in "local mode", go to /etc/pve/nodes, there should be a folder with the old node name, and most likely one with...

Was läuft in den containern? Speziell: welche systemd versionen?

And does this cause any actual problems for you?

Wie alt ist "etwas älter"? Wenn der Kernel alt genug ist würde der Bug aus dem Link eventuell noch nicht vorhanden sein.

See here[1] (The 'split image' section) for how lxd images work. You can use the vzdump tarball as a filesystem-only tarball and have to create a metadata file in an extra tarball. From what I can tell it should be enough to provide the architecture and creation_date entries (the creation date...

See https://forum.proxmox.com/threads/systemd-v239-networkd-resolved-containers-broken-by-aa.45372/#post-216864

You missed a step: Verify that the backup actually works after a restore. ;-)

Yes, this is being worked on. There will be some quirks to deal with, however. (Particularly docker and systemd-networkd manage to have conflicting requirements for when they're running in a user namespace, for example.)

There are two steps here: The first one resizes the images, this was successful. The second one was to resize the filesystem on top of that image, this failed. You'll want to shutdown the container, run fsck, mount it and then run resize2fs manually: # pct fsck 104 # pct mount 104 # resize2fs...

Depends on how long it takes. The fixes need to appear upstream first.

Apparmor in the ubuntu bionic kernel supports a new feature (socket mediation) which the userspace tools don't make use of because upstream it was moved to a newer ABI version. (Iow. our kernel doesn't advertise the feature to the user space tools and therefore they don't know about it and...

Stimmt. Da sollten wir die route gleich mit dazu eintragen, wenn es der config typ zulässt, wie bei containern.