Search results

Any way to check the (nfs) mount inside the container? Cron's @reboot does not seem to work so well inside LXC.

I just run in to this problem with the latest pfSense on latest Proxmox. WAN clients can't access any LAN Containers on the local Proxmox node. I have been scratching my head with this one for a while but then moved the LAN Containers on another Proxmox node and they started to connect with...

It's kind of scary that LXC container can just crash the host like that.

Sounds about right.

Interesting thing was that the other nested KVM Proxmox on the same harware host did not suffer any problems while that one in the picture (that held VPN LXC container) died completely.

I moved the VPN LXC container on a nested (KVM) Proxmox for better results. There were no problems for many days until last night.

I'm testing now with SMB / CIFS and it seems to work well with /var/vmail folder since Dovecot does not use symlinks but not so well with /var/www folder since ISPconfig control panel does use symlinks.

Looks like the latest kernel helped. I have not had any problems since kernel: 4.13.16-1-pve.

I'm planning to create a lxc container that uses mounted storage for /var/vmail folder. Any advice on how to do that with SMB or NFS? (on QNAP NAS).

@lao-tseu THX

Plaah. Ignore my comment about start time. That all ready exists on the "Status" page. So only my comment about "progress" ie. status is relevant.

Small request... I'dd like to see the time and progress information on the LXC "task viewer" "progress" log. When the dump starts the start time should be the first thing in the log from the start while the dump is in progress! This is especially important if we are backing up a large LXC...

I don't think it's a storage issue on our VPN server. No traffic is actually stored unless you count Swap and even that is disabled. Has to be a network traffic issue. Just out of curiosity - do you have Broadcom NIC's?

Both server hardware that I tested the VPN LXC had plenty of free resources so it's not those. Even Bandwidth was not tapping out (barely 20M/s. I'm thinking it's ones again a kernel/bridge issue with mis-handling of TX packet's. Wouldn't be the first time! The odd thing about this is that there...

Crashed again last night. Here is one of the last log entries from crashed VPN server log. I'm wondering why the client log is talking about KVM when it's on LXC container? vpn kernel: kvm: SMP vm created on host with unstable TSC; guest TSC will not be reliable

Installed the latest Linux 4.13.13-6-pve #1 SMP PVE 4.13.13-42 (Fri, 9 Mar 2018) Let's see if that keeps crashing.

Sry. Just noticed that there is newer kernel available. I had Linux - 4.13.13-6-pve #1 SMP PVE 4.13.13-41 (Wed, 21 Feb 2018) I will try the latest.

I think this is a kernel issue with LXC. Not backup. I also have had several crashes on LXC container that runs a high load VPN. Latest software and kernels. I tried to move it to different server with no change. High-Bandwidth load on LXC can crash host kernel. Now why does this sound so familiar?

Can confirm this. (Test LAN network no IPv6 DHCP server.) With LXC container configure to use DHCP with IPv6 the container just keeps trying to find IPv6 address for-everrrrr... Sam

Wiki's is 2 general. This is so specific that it can actually be converted to script by just changing the (RAID controller- and) drive names. Parted is also a light year ahead of other disk managers because it can be scripted easily.